Investigating the occurrence and fate of anticancer drugs in sewage treatment works and the wider aquatic environment

The occurrence of pharmaceuticals in wastewater and the wider environment is of growing concern. This thesis focuses on anticancer drugs - a group of biologicallypotent and often recalcitrant set of chemicals whose fate and impact on the wider freshwater environment is poorly studied. The aims of this thesis were to prioritise a group of anticancer drugs for environmental monitoring programmes (from the many drugs in use), based on their consumption and fate during wastewater treatment; to undertake a national and regional survey of two commonly used anticancer drugs, cyclophosphamide (CP) and ifosfamide (IF) in wastewater and river water; to assess the performance of a river-based chemical fate model through comparisons with field observations; and to conduct a mass balance for CP in wastewater treatment plants to assess chemical fate during the different stages of wastewater treatment. Given the large number of anticancer drugs currently in use (>70) a decision support process was developed to ascertain a short list of drugs which are most likely to persist and be released with treated effluent to environmental waters. To do this, accurate consumption data were compiled from a hospital survey in NW England and combined with urinary excretion rates derived from clinical studies. Physical– chemical property data were then compiled along with likely chemical fate and persistence during and after wastewater treatment. A shortlist of 15 chemicals (from 65), including CP and IF, was prioritised based on their consumption, persistency and likelihood of occurrence in surface waters and supported by observational studies where possible. The ecological impact of these ‘prioritised’ chemicals however is uncertain as the measured concentrations in surface waters generally fall below standard toxicity thresholds, although there is evidence that exposure of aquatic organisms to some of these chemicals may induce low-dose genotoxic effects. This prioritised sub-list of anticancer drugs should prove useful for developing environmental screening programmes and targeted toxicity assays. To assess the occurrence of anticancer drugs in wasterwaters both CP and IF were measured in raw influent and final effluent waters from fourteen STPs located across England using a sensitive analytical method. CP was detected in both wastewater influent and effluent with mean (SD) concentration of 4.1 ng/L (4.8) and 6.6 ng/L (6.5), respectively, in agreement to measured ranges from a limited number of studies conducted in Europe and elsewhere. IF was only detected in four wastewater samples with the highest concentration being observed in wastewater effluent at 0.77 ng/L (cv = 24.3% (n=3)) and possibly reflecting the relatively lower consumption of this drug relative to CP. Additional monitoring was conducted in the rivers Calder, Darwen and Ribble (North West UK) with CP present at 5 of the 6 river locations with concentrations ranging from 0.41 to 3.71 ng/L. All these rivers receive treated wastewater effluent from sewage treatment works serving different population sizes, with CP measured in river water some ~20 miles downstream of the nearest STP, indicating the widespread dispersal and persistence of this chemical. CP and IF were measured systematically down the Rivers Aire and Calder in NE England and the results compared to a GIS-based water quality model (LF2000- WQX) used to predict CP and IF distributions in the two rivers, using regional consumption data and subsequent release quantities from STPs. CP was detected in 90% of river samples, apart from rural/uplands sites located at the source of the River Aire and Calder, respectively. CP presented the highest concentration, ranging from 0.17 to 4.53 ng/L (average 1.14 ng/L). IF was seldom detected in the sampled sites and concentrations ranged from < LOD to 1.82 ng/L (average 0.51 ng/L). Model results showed a fair agreement to the measured data for CP in the River Aire, discrepancies arise as the river progressed further downstream where the modelled data was lower than the measured data. A significant input of CP from Leeds STP at A7 (STP-1) saw the continuing rise in CP despite the increase in river flow. At the lower end of the Calder (pre-confluence with the River Aire) a spike in CP is detected far beyond the modelled value. A risk assessment was carried out to establish the potential adverse effects of anticancer drugs in the river catchment. All calculated risk quotients were below 1, showing no significant risk to aquatic organisms. However, long term toxicity studies for these chemicals are needed to define the environmental stress produced by their continuous exposure and induction. The fate and removal efficiency of cyclophosphamide (CP) and ifosfamide (IF) were investigated in two conventional sewage treatment plants (STP-S and STP-C) during different stages of waste water treatment. Overall average concentrations of CP were 1.17±1 ng/L in the two plants, which is lower than recent measurements conducted elsewhere. Grab-samples were coordinated with the hydraulic residence time of wastewater in each of the treatment stages in order to monitor changes in CP concentrations in the same parcel of water as it passed through the STP. Interestingly, concentrations of CP were observed to increase from raw influent to final tertiarytreated effluent and this is likely to be attributable to the degradation of a CPmetabolite and subsequent ‘liberation’ of the parent CP as the metabolite passes through the various sewage treatment processes. This observation, apparent in both studied STPs, has implications for chemical fate modelling of anti-cancer drugs, especially if STP influent loads are used to predict subsequent fluxes to receiving waters rather than final effluent values. Moreover, this increase in concentrations made a mass balance difficult to achieve, but highlighted that elimination/removal of CP in wastewater during primary to tertiary processing is very low (