The occupational risk of environmental contamination during the storage, reconstitution, administration of antineoplastic drugs and the elimination of residues is well documented [1–4]. The chemical and physical properties of the drug, the quantity administered, the availability of personal and collective protection devices and the worker's skill determine the level of antiblastic contamination.
Several studies carried out at hospital units have shown detectable levels of cytotoxic agents in the air [5–7], on surfaces [8–15], on gloves [8, 14], and on different parts of the body [7, 8, 16]. Biological monitoring methods have been developed to detect occupational exposure to antineoplastic agents . The presence of these drugs in the urine of hospital personnel has been widely studied [7, 9, 18–20]. This has lead several organizations to develop guidelines or recommendations with the aim to improve safety during the handling of antineoplastic drugs and reduce risk of contamination in the workplace [21–23]. Based on these findings, guidelines have been also published in Italy .
Many anticancer agents have the potential to cause genetic alterations, which may lead to the development of cancer if they occur in proto-oncogenes or tumour-suppressor genes, which are involved in controlling cell growth or differentiation . Accordingly, several antineoplastic drugs have been classified by the International Agency of Research on Cancer (IARC), on the basis of epidemiological reports, animal carcinogenicity data, as well as the outcomes of in vitro genotoxicity studies, as definite (Group 1), probable (Group 2A) or possible (Group 2B) human carcinogens [26–29].
Although health care workers are exposed to much lower doses than cancer patients are, low-dose exposure over long periods can have long-term health effects.
Several epidemiological studies have been conducted investigating the cancer risks of nurses exposed to antineoplastic drugs. Increased risks for leukaemia and breast cancer were reported by Skov et al.  and Gunnarsdottir et al. . In a more recent research article, Ratner et al.  performed a cohort study among over 56,000 Canadian female nurses from British Columbia and concluded that subjects potentially exposed to antineoplastic drugs through their employment had an elevated risk of breast and rectal cancer.
Following environmental monitoring studies on contamination of workplaces from antineoplastic drugs, several biological monitoring studies have been performed. A number of studies indicate that antineoplastic drugs may cause increased genotoxic effects in pharmacists and nurses exposed in the workplace.
Undeger et al.  reported a significantly higher frequency of DNA damage - analysed using the alkaline single cell gel electrophoresis technique (comet assay) - in lymphocytes of nurses handling antiblastic drugs compared to unexposed controls; the DNA damage was, however, found to be significantly lower in nurses using compulsory personal protection equipment during their work. 8-hydroxy-2'-deoxyguanosine (8OHdG) - presumed to be an expression of oxidative damage to DNA - has never been used in assessing the mutagenic risk of occupational exposure to antineoplastic drugs.
Chromosome aberration (CA) frequencies in patients undergoing chemotherapy were significantly higher than in controls [34–37]. Increased CA frequencies have also been found in hospital personnel handling cytotoxic drugs [18, 38–43]. Negative findings have also been reported, however [44–47]. In hospitals where nurses used inadequate safety cabinets when handling cytostatics, significantly elevated levels of CAs (as well as sister chromatid exchanges, SCEs, and unscheduled DNA-repair synthesis) were detected by Jakab et al. .
According to Kevekordes et al. , a malfunctioning safety hood resulted in a higher frequency of micronuclei (MN) and SCE in exposed nurses compared to matched controls. Kasuba et al.  found that the length of handling cytostatic drugs increased the frequency of MN, whereas no statistically significant difference was observed for SCE. In hospital pharmacy personnel adopting high standards of safety, Pilger et al.  observed that frequencies of MN and SCE were similar to those of controls, whereas small increases in these genetic end-points were found in accidental contamination events. Anwar et al.  found statistically significant increases in both CAs and MN in nurses handling cytostatic drugs. Hessel et al.  reported no association between antiblastic drugs in urine and MN frequency in lymphocytes of exposed hospital workers. Lastly, Maluf & Erdtmann , when comparing pharmacists and nurses exposed to antineoplastic drugs with unexposed controls, found no statistical difference for MN and dicentric bridge frequency, whereas the mean value of DNA migration detected by comet assay was significantly higher in the exposed group compared to the controls.
Several studies showed the influence of metabolic and DNA repair polymorphisms on biological indicators of genotoxic risk (urinary metabolites, protein and DNA adducts, citogenetic tests), which are commonly used in the biomonitoring of occupational exposure to antineoplastic agents . There are however few studies on the influence of genetic polymorphisms of enzymes involved in DNA damage induced by occupational exposure to antineoplastic drugs [55, 56].
In conclusion, some chemical studies have assessed occupational exposure to antineoplastic drugs, and some epidemiological investigations have detected various toxicologic effects in groups of exposed workers. To our knowledge, only very few researches have studied the same population by simultaneous assessment of exposure, biologic effects and genetic susceptibility .
The protocol of this molecular epidemiology study therefore presents an integrated chemical and biotoxicological approach for environmental and biological monitoring of exposure and cancer risks which will be implemented in a large number of healthy non-smoking female hospital nurses. This approach is based on monitoring procedures reported on the Italian guidelines  which includes, beside methods for preventing exposure to antineoplastic drugs, also monitoring recommendations. In particular, the guidelines provides guidance on the control of antineoplastic drug contaminations on surfaces and clothes by environmental monitoring (wipe and pad tests, respectively) and on the control of exposure by biological monitoring (concentrations of antineoplastic drugs in body fluids, usually urine), with both contamination and exposure depending on working practices and the frequency and adequacy of decontamination procedures. In this context, the advantage of biological monitoring is being able to measure the total uptake of antineoplastic drugs by all routes of exposure, however, testing is generally limited to one or very few agents that are considered as model compounds. In this study, monitoring of genotoxic risks we will performed by combining environmental and biological monitoring as above, with procedures for biological effect monitoring (urinary 8OHdG, DNA strand breakage and cytogenetic abnormalities in lymphocytes) as well as genetic susceptibility monitoring (metabolic polymorphisms).