A practical guide to epidemiological practice and standards in the identification and validation of diagnostic markers using a bladder cancer example.

Technical advances to analyze biological markers have generated a plethora of promising new marker candidates for early detection of cancer. However, in subsequent analyses only few could be successfully validated as being predictive, clinically useful, or effective. This failure is partially due to rapid publication of results that were detected in early stages of biomarker research. Methodological considerations are a major concern when carrying out molecular epidemiological studies of diagnostic markers to avoid errors that increase the potential for bias. Although guidelines for conducting studies and reporting of results have been published to improve the quality of marker studies, their planning and execution still need to be improved. We will discuss different sources of bias in study design, handling of specimens, and statistical analysis to illustrate possible pitfalls associated with marker research, and present legal, ethical, and technical considerations associated with storage and handling of specimens. This article presents a guide to epidemiological standards in marker research using bladder cancer as an example. Because of the possibility to detect early cancer stages due to leakage of molecular markers from the target organ or exfoliation of tumor cells into the urine, bladder cancer is particularly useful to study diagnostic markers. To improve the overall quality of marker research, future developments should focus on networks of studies and tissue banks according to uniform legal, ethical, methodological, and technical standards. This article is part of a Special Issue entitled: Computational Proteomics in the Post-Identification Era. Guest Editors: Martin Eisenacher and Christian Stephan.

Biomarker research with prospective study designs for the early detection of cancer.

This article describes the principles of marker research with prospective studies along with examples for diagnostic tumor markers. A plethora of biomarkers have been claimed as useful for the early detection of cancer. However, disappointingly few biomarkers were approved for the detection of unrecognized disease, and even approved markers may lack a sound validation phase. Prospective studies aimed at the early detection of cancer are costly and long-lasting and therefore the bottleneck in marker research. They enroll a large number of clinically asymptomatic subjects and follow-up on incident cases. As invasive procedures cannot be applied to collect tissue samples from the target organ, biomarkers can only be determined in easily accessible body fluids. Marker levels increase during cancer development, with samples collected closer to the occurrence of symptoms or a clinical diagnosis being more informative than earlier samples. Only prospective designs allow the serial collection of pre-diagnostic samples. Their storage in a biobank upgrades cohort studies to serve for both, marker discovery and validation. Population-based cohort studies, which may collect a wealth of data, are commonly conducted with just one baseline investigation lacking serial samples. However, they can provide valuable information about factors that influence the marker level. Screening programs can be employed to archive serial samples but require significant efforts to collect samples and auxiliary data for marker research. Randomized controlled trials have the highest level of evidence in assessing a biomarker's benefit against usual care and present the most stringent design for the validation of promising markers as well as for the discovery of new markers. In summary, all kinds of prospective studies can benefit from a biobank as they can serve as a platform for biomarker research. This article is part of a Special Issue entitled: Biomarkers: A Proteomic Challenge.

Effects of Exposure to Welding Fume on Lung Function: Results from the German WELDOX Study.

The association between exposure to welding fume and chronic obstructive pulmonary disease (COPD) has been insufficiently clarified. In this study we assessed the influence of exposure to welding fume on lung function parameters. We investigated forced expiratory volume in 1 s (FEV1), forced vital capacity (FVC), FEV1/FVC, and expiratory flow rates in 219 welders. We measured current exposure to respirable particles and estimated a worker's lifetime exposure considering welding techniques, working conditions and protective measures at current and former workplaces. Multiple regression models were applied to estimate the influence of exposure to welding fume, age, and smoking on lung function. We additionally investigated the duration of working as a welder and the predominant welding technique. The findings were that age- and smoking-adjusted lung function parameters showed no decline with increasing duration, current exposure level, and lifetime exposure to welding fume. However, 15% of the welders had FEV1/FVC below the lower limit of normal, but we could not substantiate the presence of an association with the measures of exposure. Adverse effects of cigarette smoking were confirmed. In conclusion, the study did not support the notion of a possible detrimental effect of exposure to welding fume on lung function in welders.

New malignancies following cancer of the urinary bladder: analysis of German cancer registry data.

This analysis aimed at occurrence and distribution patterns of new malignancies following bladder cancer. Standardised incidence ratios (SIRs) were calculated for two German population-based cancer registries of North Rhine-Westphalia (NRW) and Saarland to access risks for subsequent primaries. An elevated risk for secondary cancer of any site but urothelium was observed in NRW men [SIR 1.35, 95% confidence interval (CI) 1.22-1.49]. The corresponding risk in Saarland was not significantly elevated (SIR 1.06, 95% CI 0.97-1.15). In data of both registries excess risks were observed for cancer of the respiratory tract (SIR 1.54, CI 1.23-1.89 in NRW men) and the prostate (SIR 1.91, 95% CI 1.61-2.24 in NRW; SIR 1.25, 95% CI 1.07-1.45 in Saarland). Common risk factors and incidental findings during follow-up care of bladder cancer patients might explain most of the observed patterns. In addition SIRs were throughout particular high for subsequent cancer of the renal pelvis and the ureter due to pathological characteristics of urothelial neoplasms.

Performance of biomarkers SMRP, CA125, and CYFRA 21-1 as potential tumor markers for malignant mesothelioma and lung cancer in a cohort of workers formerly exposed to asbestos.

The aim of the study is to examine the cancer-predictive values of SMRP (soluble mesothelin-related peptides), CA125, and CYFRA21-1 as potential tumor markers for lung cancer and malignant mesothelioma in a cohort of workers formerly exposed to asbestos. A voluntary surveillance program has been established for German workers with former asbestos exposure. A subgroup of 626 subjects with a mean age of 63 years (range 53-70 years) at baseline was enrolled in an extended health examination program with high-resolution computer tomography (HRCT) of the chest and blood drawing between 1993 and 1997. Serum concentrations of SMRP, CA125, and CYFRA21-1 were measured in archived serum samples in 2005 and 2006. A mortality follow-up was conducted through 2007. So far, 12 cases with lung cancer and 20 cases with malignant mesothelioma have been observed in this cohort. The average time between sample collection and diagnosis was 4.7 years. Analyzed biomarkers showed low sensitivities (5-25%) and positive predictive values (4-30%) for both cancer sites. Marker combinations resulted in sensitivities between 5 and 50% and positive predictive values ranging from 3 to 14%. Even in those cases, where biomarker concentrations were available within 36 months before diagnosis, no trend for increasing biomarker levels was observed. The analyzed tumor markers were characterized by high specificities, but low sensitivities. SMRP, CA125, and CYFRA21-1 alone or in combination were less suitable to serve as predictors for the diagnosis of lung cancer or malignant mesothelioma. However, a prospective study with annual sampling might reveal a better predictive value of these markers.

Occupational risks for adenocarcinoma of the nasal cavity and paranasal sinuses in the German wood industry.

OBJECTIVES: To examine the risk of wood dust and chemical exposures for adenocarcinoma of the nasal cavity and paranasal sinuses (ADCN) among German wood workers. METHODS: An industry-based case-control study with 86 male ADCN cases and 204 controls was conducted in the German wood-working industries. Cumulative and average wood-dust exposure was quantified with a job-exposure matrix based on wood-dust measurements at recent and historical workplaces. Probabilities of exposure to wood preservatives, stains, varnishes, and formaldehyde were semi-quantitatively rated. Odds ratios and 95% confidence intervals were calculated with logistic regression analysis conditional on age and adjusted for smoking and other factors. For estimating the risks of either wood dust or chemical additives, the authors additionally adjusted for the corresponding co-exposure. RESULTS: ADCN occurred relatively more frequently among wood workers that had ever worked as cabinet makers or joiners (OR 2.96, 95% CI 1.46 to 6.01) than as saw millers (OR 0.15, 95% CI 0.03 to 0.68). Average exposure to inhalable wood dust >/=5 mg/m(3) was associated with a high risk (OR 48.47, 95% CI 13.30 to 176.63) compared to levels below 3.5 mg/m(3). Assuming 40 years of exposure under these concentrations, the corresponding OR was 4.20 (95% CI 1.69 to 10.43). Exposure between 3.5 and 5 mg/m(3) was also found to pose a risk (OR 10.54, 95% CI 3.34 to 33.27). Exposure to pigment stains before 1970 was associated with an increased risk (OR 3.03; 95% CI 1.11 to 8.26). No significant associations were estimated for wood preservatives, varnishes, and formaldehyde. CONCLUSIONS: The authors found an elevated ADCN risk for exposure to inhalable wood dust above 3.5 mg/m(3). The rareness of the disease does not allow the exclusion of risk below that concentration. For pigment stains, there is evidence for an association of historical exposure with the development of ADCN in German wood workers.

Exhaled breath condensate analysis: evaluation of a methodological setting for epidemiological field studies.

The analysis of exhaled breath condensate (EBC) is a noninvasive method to determine airway inflammation. There is accumulating evidence that pH, nitrogen oxides (NOx), and leukotriene B4 (LTB4) are suitable markers of inflammatory airway diseases. It was the aim of the present pilot study to evaluate a protocol of EBC sampling and processing for epidemiologic field studies. EBC samples were repeatedly obtained by 16 healthy adults on each of four days over two consecutive weeks. The effects of sample collection (volume vs. time, with vs. without nose-clip) and sample processing on volume, pH, NOx, and LTB4 as well as within-day, between-day, and between-week reproducibility were assessed. We found that wearing a nose-clip had no significant effects on mediators. Sampling time was the major determinant of the volume. Reproducibility was fairly good. Storage at 4-6 degrees C for 24 h had no significant effects on NOx and LTB4 concentrations, but resulted in an increase of pH. In conclusion, wearing a nose-clip during EBC collection is not recommended in terms of convenience and compliance for repeated investigations. Samples for the analyses of NOx and LTB4 can be transported under usual conditions, but pH should preferably be determined at the workplace.

Dispersion of As and selected heavy metals around a coal-burning power station in central Slovakia.

A power station in central Slovakia emitted arsenic (As) in large quantities for over 30 years as a result of burning As-rich brown coal. Nowadays emissions of As are low. Over the lifetime of the plant's operation over 3000 tonne of As have been emitted into the environment. This paper aims to examine the concentrations of As in the soil around the power station, and also to investigate whether the coal burnt in the plant, and consequently the emissions from it, contained raised levels of six further heavy metals. Soil concentrations were compared to ground level air As concentrations predicted by an air dispersion model. Coal samples were taken from the power station and analysed to determine concentrations of As, Zn, Pb, Cu, Cr, Ni and Cd. Soil samples (n=113) were taken up to 12 km from the plant along a transect designed to follow the valley floor in which the power station is situated. Soil samples were analysed for concentrations of those elements for which coal was tested. Concentrations of As in coal were high (AM 518 mug/g). Those of other heavy metals were, in general, low. Concentrations of soil As were substantially raised in the near vicinity of the plant but decreased within 5 km to concentrations similar to those in the rest of the district. Overall, levels within 10 km of the plant were slightly above those recommended for residential levels in the UK. Soil concentrations of other heavy metals were higher in the vicinity of the plant but none, overall was raised. Comparison of results from a previous air dispersion model of ground level air arsenic concentrations showed a moderate correlation (r=0.6) between modelled and measured values. Over its period of operation the power plant has contributed to raised levels of soil As in the local soils, though not substantially of other elements. Though now airborne As emissions are controlled, concern remains regarding soil arsenic concentrations and fugitive emissions from the plant that could be contributing to exposure of the local population and of the workforce.

A simple and rapid flow cytometric method for detection of porcine cell surface markers.

The objective of this study was to develop a rapid and reliable method for flow cytometric analysis of porcine whole blood cells. Fifty-microliters of heparin- or EDTA-treated whole blood was added to wells of a round-bottom 96-well microtitration plate. Each well contained 10 microl of an appropriate dilution of four different antibodies (40 microl total; two primary monoclonal antibodies and two fluorescent-labeled secondary antibodies). For convenience, the antibody mixture could be added to plates 1-2 days prior to assay and stored at 4 degrees C. Once whole blood was added to wells, plates were mixed gently, placed in a sealed bag and incubated in the dark at room temperature for 20 min. Contents of wells were then transferred to polystyrene tubes containing 2 ml of 1.5% formalin in distilled water and mixed gently. Cells were fixed for a minimum of 30 min and then stored in the dark at 4 degrees C until analysis by flow cytometry. Analysis of cell samples may be done up to 3 days after fixation. Results indicate that the percentages of Class I, Class II, CD3, CD8, CD4, CD45, monocyte, gamma-delta T-cell populations, and total number of granulocytes identified using this method were comparable to standard values or to values obtained following separation of white blood cells from red blood cells. The percentage of labeled B-cells was lower than standard values. Total assay time from receipt of blood to acquisition of data by flow cytometry required less than 2 h. This modified assay was shown to be simple, reliable, and useful for screening large numbers of porcine samples in a minimal period of time.

Occupational risk factors for renal cell carcinoma: agent-specific results from a case-control study in Germany. MURC Study Group. Multicenter urothelial and renal cancer study.

BACKGROUND: This case-control study was conducted to estimate the renal cell cancer (RCC) risk for exposure to occupation-related agents, besides other suspected risk factors. METHODS: In a population-based multicentre study, 935 incident RCC cases and 4298 controls matched for region, sex, and age were interviewed between 1991 and 1995 for their occupational history and lifestyle habits. Agent-specific exposure was expert-rated with two job-exposure matrices and a job task-exposure matrix. Conditional logistic regression was used to calculate smoking adjusted odds ratios (OR). RESULTS: Very long exposures in the chemical, rubber, and printing industries were associated with risk for RCC. Males considered as 'substantially exposed to organic solvents' showed a significant excess risk (OR = 1.6, 95% CI : 1.1-2.3). In females substantial exposure to solvents was also a significant risk factor (OR = 2.1, 95% CI : 1.0-4.4). Excess risks were shown for high exposure to cadmium (OR = 1.4, 95% CI : 1.1-1.8, in men, OR = 2.5, 95% CI : 1.2-5.3 in women), for substantial exposure to lead (OR = 1.5, 95% CI : 1.0-2.3, in men, OR = 2.6, 95% CI : 1.2-5.5, in women) and to solder fumes (OR = 1.5, 95% CI : 1.0-2.4, in men). In females, an excess risk for the task 'soldering, welding, milling' was found (OR = 3.0, 95% CI : 1.1-7.8). Exposure to paints, mineral oils, cutting fluids, benzene, polycyclic aromatic hydrocarbons, and asbestos showed an association with RCC development. CONCLUSIONS: Our results indicate that substantial exposure to metals and solvents may be nephrocarcinogenic. There is evidence for a gender-specific susceptibility of the kidneys.

Occupational risk factors for urothelial carcinoma: agent-specific results from a case-control study in Germany. MURC Study Group. Multicenter Urothelial and Renal Cancer.

BACKGROUND: This multicentre population-based case-control study was conducted to estimate the urothelial cancer risk for occupational exposure to aromatic amines, polycyclic aromatic hydrocarbons (PAH), and chlorinated hydrocarbons besides other suspected risk factors. METHODS: In a population-based multicentre study, 1035 incident urothelial cancer cases and 4298 controls matched for region, sex, and age were interviewed between 1991 and 1995 for their occupational history and lifestyle habits. Exposure to the agents under study was self-assessed as well as expert-rated with two job-exposure matrices and a job task-exposure matrix. Conditional logistic regression was used to calculate smoking adjusted odds ratios (OR) and to control for study centre and age. RESULTS: Urothelial cancer risk following exposure to aromatic amines was only slightly elevated. Among males, substantial exposures to PAH as well as to chlorinated solvents and their corresponding occupational settings were associated with significantly elevated risks after adjustment for smoking (PAH exposure, assessed with a job-exposure matrix: OR = 1.6, 95% CI: 1.1-2.3, exposure to chlorinated solvents, assessed with a job task-exposure matrix: OR = 1.8, 95% CI: 1.2-2.6). Metal degreasing showed an elevated urothelial cancer risk among males (OR = 2.3, 95% CI: 1.4-3.8). In females also, exposure to chlorinated solvents indicated a urothelial cancer risk. Because of small numbers the risk evaluation for females should be treated with caution. CONCLUSIONS: Occupational exposure to aromatic amines could not be shown to be as strong a risk factor for urothelial carcinomas as in the past. A possible explanation for this finding is the reduction in exposure over the last 50 years. Our results strengthen the evidence that PAH may have a carcinogenic potential for the urothelium. Furthermore, our results indicate a urothelial cancer risk for the use of chlorinated solvents.

Challenges to environmental toxicology and epidemiology: where do we stand and which way do we go?

Modern toxicology investigates a wide array of both old and new health hazards. Priority setting is needed to select agents for research from the plethora of exposure circumstances. The changing societies and a growing fraction of the aged have to be taken into consideration. A precise exposure assessment is of importance for risk estimation and regulation. Toxicology contributes to the exploration of pathomechanisms to specify the exposure metrics for risk estimation. Combined effects of co-existing agents are not yet sufficiently understood. Animal experiments allow a separate administration of agents which can not be disentangled by epidemiological means, but their value is limited for low exposure levels in many of today's settings. As an experimental science, toxicology has to keep pace with the rapidly growing knowledge about the language of the genome and the changing paradigms in cancer development. During the pioneer era of assembling a working draft of the human genome, toxicogenomics has been developed. Gene and pathway complexity have to be considered when investigating gene-environment interactions. For a best conduct of studies, modern toxicology needs a close liaison with many other disciplines like epidemiology and bioinformatics.

Lymphocyte recirculation in cattle: patterns of localization by mammary and mesenteric lymph node lymphocytes.

We examined patterns of lymphocyte localization in female dairy cattle following infusion of 51Cr-labeled autologous lymphocytes prepared from surgically excised mammary or ileal mesenteric lymph nodes. Labeled lymphocytes prepared from mammary lymph nodes were recovered in proportionally high numbers from mammary and prescapular lymph nodes, and in low numbers from intestinal mesenteric nodes. This pattern was observed in both heifers and lactating cows. In contrast, labeled lymphocytes prepared from ileal mesenteric lymph nodes of lactating cows were recovered in proportionally high numbers from intestinal mesenteric nodes, and in low numbers from mammary and prescapular nodes. These findings, when compared with previous results in sheep and swine, support the hypothesis that lymphocytes do not migrate efficiently between the gut and mammary gland of ruminants.

Extravasation of lymphocytes via paracortical venules in sheep lymph nodes: visualization using an intracellular fluorescent label.

In rodents and humans, lymphocytes extravasate into lymph nodes via specialized paracortical venules lined with high endothelium (HEV). Sheep and other ruminants do not have morphologically defined HEV in their lymph nodes. It has been assumed that lymphocyte extravasation in these species proceeds via analogous structures; i.e., paracortical venules lined with low to medium endothelium. In this study, lymphocyte suspensions were prepared from surgically excised lymph nodes of sheep and labeled with an intracellular fluorescent dye, H33342. Labeled cells were infused intravenously back into donors, and sheep were killed at various intervals after infusion. Frozen sections of lymph nodes were examined microscopically for the location of labeled cells. Ten minutes after infusion, labeled cells were seen in the lumen of venules located in the paracortical region of the nodes. At later time points, cells were seen apparently migrating through the venule walls and in the adjacent paracortical tissue. Similar experiments were performed in which H33342-labeled murine lymphocytes were infused into syngeneic mice. When equivalent cell numbers (based on animal size) were infused, no obvious differences were seen between location and kinetics of appearance of labeled cells in lymph nodes of sheep compared to those of mice. These results indicate that lymphocyte extravasation in sheep proceeds via paracortical venules in lymph nodes. The function of these venules appears to be analogous to HEV in nonruminant species.

Characterization of mitogen-stimulated porcine lymphocytes using a stable fluorescent dye (PKH2) and multicolor flow cytometry.

Stimulation of lymphocyte proliferation using mitogens or specific antigens is a method that is used frequently to assess immune responsiveness. While useful, lymphocyte blastogenesis, or [3H]-thymidine incorporation, provides little information regarding the response of specific subsets to the stimulant. Here, we report that the fluorescent cell membrane probe, PKH2, is a useful tool for measuring the proliferation of porcine lymphocyte subpopulations by utilizing multicolor flow cytometry. For this study, mitogen-induced proliferation of porcine peripheral blood mononuclear cells (PBMCs) was measured using [3H]-thymidine incorporation as well as a flow cytometric-based proliferation assay. From the [3H]-thymidine incorporation data alone, it was observed that PBMC stimulated with either concanavalin A (Con A), phytohemagglutinin (PHA) or pokeweed mitogen (PWM) demonstrated greater proliferation on day 3 than on day 5 of culture. Using the PKH dye and flow cytometric analysis, the responsiveness of specific lymphocyte subsets to mitogen stimulation was detected. The predominant subsets of porcine lymphocytes responding to Con A or PHA stimulation were CD4(+)CD8(+), CD4(-)CD8alpha(hi), CD4(-)CD8alpha(lo) and gammadelta TCR(+) cells. PWM stimulation induced responses by CD4(+)CD8(+), CD4CD8alpha(hi) but not by CD4(-)CD8alpha(lo) or gammadelta TCR(+) cells. Con A stimulation resulted in a sustained proliferation of CD8alpha(hi) cells over the 5-day period while PHA stimulation resulted in proliferation that peaked within the first 3 days. Little or no proliferative responses were detected within the IgM(+) population (e.g. B cells). This is the first study to define the contribution of individual lymphocyte subsets to mitogen-induced proliferation of porcine PBMCs.

MHC class II-restricted, CD4(+) T-cell proliferative responses of peripheral blood mononuclear cells from Mycobacterium bovis-infected white-tailed deer.

White-tailed deer are significant wildlife reservoirs of Mycobacterium bovis for cattle, predators, and, potentially, humans. Infection of cattle with M. bovis stimulates an antigen-specific T-cell response, with both CD4(+) and CD8(+) cells implicated in protective immunity. Few studies, however, have examined lymphocyte subset responses to experimental M. bovis infection of white-tailed deer. In this study, a flow cytometric proliferation assay was used to determine the relative contribution of individual peripheral blood mononuclear cell subsets of M. bovis-infected white-tailed deer in the recall response to M. bovis antigen. Naive deer were challenged with M. bovis by cohabitation with infected deer. These M. bovis-challenged deer developed significant in vivo (delayed-type hypersensitivity) and in vitro (proliferative) responses to M. bovis purified protein derivative (PPD). At necropsy, typical tuberculous lesions containing M. bovis were detected within lungs and lung-associated lymph nodes of infected deer. The predominant subset of lymphocytes that proliferated in response to in vitro stimulation with PPD was the CD4(+) subset. Minimal proliferative responses were detected from CD8(+), gamma delta TCR(+), and B-cells. Addition of monoclonal antibodies specific for MHC II antigens, but not MHC I or CD1 antigens, abrogated the proliferative response. Together, these findings indicate that while CD4(+) cells from infected deer proliferate in the recall response to M. bovis antigens, this response is not sufficient to clear M. bovis and immunologic intervention may require stimulation of alternate subsets of lymphocytes.

Lymphocyte subset proliferative responses of Mycobacterium bovis-infected cattle to purified protein derivative.

Despite highly successful eradication efforts in several countries, Mycobacterium bovis infection of cattle remains a significant health concern worldwide. Immune mechanisms of resistance to and/or clearance of M. bovis infection of cattle, however, are unclear. Recent studies have provided evidence supporting a role for CD4(+), CD8(+), and gammadelta TCR(+) T cells in the response of cattle to M. bovis. In the present study, we utilized a flow cytometric-based proliferation assay to determine the relative contribution of individual lymphocyte subsets in the response to M. bovis infection and/or sensitization with mycobacterial purified protein derivative (PPD). Peripheral blood mononuclear cells (PBMC) from M. bovis-infected cattle proliferated in response to in vitro stimulation with M. bovis PPD. CD4(+) T cells and gammadelta TCR(+) cells were the predominate subsets of lymphocytes responding to PPD. gammadelta TCR(+) cells also proliferated in non-stimulated cultures; however, the gammadelta TCR(+) cell proliferative response of infected cattle was significantly (p<0.05) greater in PPD-stimulated cultures as compared to non-stimulated cultures. Intradermal injection of PPD for comparative cervical testing (CCT) induced a boost in the in vitro proliferative response of CD4(+) but not gammadelta TCR(+) cells of infected cattle. Administration of PPD for CCT also boosted interferon-gamma (IFN-gamma) production by PBMC of infected cattle following in vitro stimulation with M. bovis PPD. Injection of PPD for CCT did not, however, elicit a proliferative or IFN-gamma response in cells isolated from non-infected cattle. These data indicate that CD4(+) and gammadelta TCR(+) cells of M. bovis-infected cattle proliferate in a recall response to M. bovis PPD and that the CD4(+) cell response is boosted by intradermal injection with PPD for CCT.

Assessment of environmental arsenic levels in Prievidza district.

A coal-burning power station in the Nitra Valley in central Slovakia annually emitted large quantities of arsenic (up to 200 tonnes) between 1953 and 1989. Since then, pollution-control measures have reduced arsenic emissions to less than 2 tonnes a year. However, the power station was still a source of airborne arsenic pollution. As part of an EU-funded study on exposure to arsenic and cancer risk in central and Eastern Europe we carried out a study of environmental levels of arsenic in the homes and gardens of residents of the district. Garden soil samples (n=210), house dust samples (n=210) and composite house dust samples (n=109) were collected and analysed using inductively coupled plasma atomic absorption spectroscopy (ICP-AES) at Imperial College. The mean arsenic content of coal and ash in samples taken from the plant was 519 microg/g (n=19) and 863 microg/g (n=22), respectively. The geometric mean (GM) arsenic concentration of garden soils was 26 microg/g (range 8.8-139.0 microg/g), for house dust 11.6 microg/g (range 2.1-170 microg/g) and for composite house dust 9.4 microg/g (range 2.3-61.5 microg/g). The correlation between the arsenic levels in soil and in house dust was 0.3 (P<0.01), in soil and composite house dust 0.4 and house dust and composite house dust 0.4 (P<0.01 for both), i.e., were moderate. Arsenic levels in both house dust and soil decreased with distance from the power station. Overall, levels in both fell by half 5 km from the point source. Weak correlations were seen between the total urinary arsenic concentrations and arsenic concentrations in composite house dust.