The potential ecotoxicological impact of pharmaceutical and personal care products on humans and freshwater, based on USEtoxTM characterization factors. A Spanish case study of toxicity impact scores.

The potential ecotoxicological impact of pharmaceutical and personal care products on humans and freshwater, based on USEtoxTM characterization factors. A Spanish case study of toxicity impact scores. Sheyla Ortiz de Garcíaa,c*, Pedro A. García-Encinaa, Rubén Irusta-Matab a) Department of Chemical Engineering and Environmental Technology, University of Valladolid, Calle Dr. Mergelina s/n, 47011, Valladolid, Spain; Phone: +34983423171; Fax: +34983423013. E-mail address: [email protected]; p[email protected]. b) Department of Chemical Engineering and Environmental Technology, University of Valladolid, Paseo del Cauce 59, 47011, Valladolid, Spain; Phone: +34983423693; Fax: +34983423310. E-mail address: [email protected]. c) Department of Chemistry, Faculty of Sciences and Technology, University of Carabobo, Av. Salvador Allende, Campus Bárbula, Carabobo State, Bolivarian Republic of Venezuela; Phone: +582418688229; Fax: +582418688229. E-mail address: [email protected].ve. * Corresponding author: Tel.: +34 983 423 171; Fax: +34 983 423 013. E-mail address: shey[email protected].es [email protected] (S. Ortiz). Abstract Pharmaceutical and personal care products (PPCPs) are being increasingly included in Life Cycle Assessment studies (LCAs) since they have brought into evidence both human and ecological adverse effects due to their presence in different environmental compartments, wastewater facilities and industry. Therefore, the main goal of this research was to estimate the characterization factors (CFs) of 27 PPCPs widely used worldwide in order to incorporate their values into Life Cycle Impact Assessment studies (LCIA) or to generate a toxicity impact score ranking. Physicochemical properties, degradation rates, bioaccumulation, ecotoxicity and human health effects were collected from experimental data, recognized databases or estimated using EPI SuiteTM and the USEtoxTM software, and were subsequently used for estimating CFs. In addition, a Spanish toxicity impact score ranking was carried out for 49 PPCPs using the 27 newly calculated CFs, and 22 CFs already available in the literature, besides the data related to the occurrence of PPCPs in the environment in Spain. It has been highlighted that emissions into the continental freshwater compartment showed the highest CFs values for human effects (ranging from 10-9 to 10-3 Cases·kg-1), followed by emissions into the air (10-9 to 10-5 Cases·kg-1), soil (10-11 to 105 Cases·kg-1) and seawater (10-12 to 10-4 Cases·kg-1). CFs regarding the affectation of freshwater aquatic environments were the highest of those proceeding from emissions into continental freshwater (between 1 to 104 PAF·m3·day·kgemission-1) due to the direct contact between the source of emission and the compartment affected, followed by soil (among 10-1 to 104 PAF·m3·day·kgemission-1), and air (among 10-2 to 104 PAF·m3·day·kgemission-1) while the lowest were the CFs of continental seawater (among 10-28 to 10-3 PAF·m3·day·kgemission-1). Freshwater aquatic ecotoxicological CFs are much higher than human toxicity CFs, demonstrating that the ecological impact of PPCPs in aquatic environments must be a matter of urgent attention. According to the Spanish toxicity impact score calculated, the PPCPs with the highest impact are hormones, antidepressants, fragrances, antibiotics, angiotensin receptor blockers and blood lipid regulators, which have already been found in other kinds of score rankings. These results, which were not available until now, will be useful in order to perform better LCIA studies, incorporating the micro-pollutants whose CFs have been estimated, or in order to carry out single hazard/risk environmental impact assessments. Keywords: Characterization Factor, Ecotoxicity, Human toxicity, Life Cycle Impact Assessment, Pharmaceuticals and Personal Care Products. 1. Introduction In recent years, pharmaceutical and personal care products (PPCPs) have been found at different levels of concentration in all environmental compartments (air, water and soil), and many of their impacts are still unknown or under analysis. The primary routes of pharmaceuticals into the environment are through human excretion, disposal of unused products and through agricultural usage, but high concentrations of pharmaceuticals have also been reported in treated industrial effluents or recipient waters, through direct discharge from manufacturing companies (Larsson et al. 2007; Larsson, 2014a). A wide range of pharmaceutical products have been detected in surface and groundwater, associated with wastewater disposal (Brausch and Rand, 2011; Ebele et al. 2017; Stuart et al., 2012) but can also be found in soil, sediments (Larsson, 2014b; Xu et al. 2009) and to a lesser extent possibly in the air (Larsson, 2014b). There is a lot of research in this area, some of which was carried out by Ebele et al. (2017) who presented a review of the current state-of-knowledge on PPCPs in the freshwater aquatic environments (water, sediments and biota) of the five continents. Wu et al. (2009) studied the degradation and adsorption of selected PPCPs in agricultural soils, while Gaw et al. (2014) reviewed the sources, impacts and concentrations of pharmaceuticals in marine and coastal environments, among many others. In the specific case of personal care products (PCPs), Tolls et al. (2009) indicated that considerable amounts of these compounds are utilized each day, resulting in large quantities of chemical substances that could potentially reach environmental compartments, particularly water, but also soil and air. Currently, there are thousands of PPCPs which are available on the market and are used daily, and they can be released into the environment individually, although in most cases they tend to be released in ever-changing mixtures, whose effects can be synergetic or antagonistic, so the possibility of knowing the potential impact that all these compounds and their mixtures might generate in nature is almost impossible, without spending a large amount of money, resources and time. The effects of PPCPs in the environment are very diverse, these compounds can be persistent, bioaccumulable and can cause acute and chronic human and ecotoxicological damage. Therefore, the interest in knowing the ecotoxicological effects of PPCPs on the environment has increased in the last years. Some authors (Brausch and Rand, 2011; Cleuvers, 2003, 2004; Daughton and Brooks, 2010; Fent et al., 2006; Sanderson et al., 2004a; Santos et al., 2010; Vasquez and Fatta- Kassinos, 2013) have reported ecotoxicogical data from different types of assays (for single PPCP or their mixtures) including different species (trophic levels), times of exposure (chronic, subchronic or acute) and endpoints (Half maximal effective concentration, EC50; concentration which causes the death of 50% of the sample population, LC50; Non observed adverse effect level, NOAEL; Lowest observed adverse effect level, LOAEL). Although there is a great variety of analyses and tests to bring into evidence the negative effect of PPCPs on the environment, it is necessary to have different methods of predicting these effects, due to the large number of these types of compounds, and the diverse ways that they can be found in the environment. A large number of tools for predicting the impact of a process, activity or contaminant in the environment are currently available. One of them is the Life Cycle Assessment (LCA), which allows the estimation of the potential impacts of such compounds on human health, ecosystems and resources. LCA has been extended to many aspects of production and consumption, including eco-design of products, cleaner production, environment labels, green purchase, resource management, wastes management and environment strategy, etc. (Nie et al., 2010), and therefore, LCA is gradually gaining acceptance as an efficient tool for the environmental evaluation of the potential impact of chemicals and chemical processes. LCA does not substitute other methodologies (such as environmental risk assessment, or the ratio between predicted environmental concentration and the predicted no effect concentration etc.) since the different tools fulfill different purposes and they can, in fact, play complementary roles and benefit from each other (Muñoz et al., 2008). Kobayashi et al. (2015) proposed the combination of LCA and quantitative risk assessment (QRA) with different hybridization approaches, taking into account that LCA is useful in the evaluation of global impacts and QRA in local impacts. The guidelines of LCA studies are established in the standard series of ISO 14040, and more specifically in the ISO 14040:2006 and 14044:2006 standards (ISO, 2006). LCA methodology can be a powerful tool: (i) to identify the type of impact (on renewable or non-renewable resources, global warming, ozone depletion, toxicity, acidification, energy and water use, among others) of diverse compounds in different environmental scenarios; (ii) to compare these impacts with those from other compounds; (iii) to implement preventive or corrective actions to minimize the potential or real adverse effect caused by them. The life cycle impact assessment (LCIA) phase of a LCA study requires not only the data from the emissions inventory, but also the characterization factors (CFs, alternatively referred to as equivalency factors) to provide indicators in the context of various impact categories (such as global warming, stratospheric ozone depletion, tropospheric ozone creation, eutrophication/nitrification, acidification, toxicological impacts on humans, and toxicological impacts on ecosystems) (Pennington et al., 2004). Knowledge of CFs is mainly essential for determining/estimating the human and ecological potential impact of chemicals on different environments (air, freshwater, seawater, natural soil, agricultural soil, etc.) and they must be included in the LCIA stage of LCA studies. CFs are also used to determine the relative importance of a substance to toxicity related impact categories, such as human toxicity or ecotoxicity in LCA studies. The CFs accounts for the environmental persistence (fate), accumulation in the human food chain (exposure) and toxicity (effect) of a chemical. Fate and exposure factors can be calculated by means of “evaluative” multimedia fate and exposure models, while effect factors can be derived from ecotoxicity data on human beings and laboratory organisms (Huijbregts et al. 2005a). In this sense, the USEtoxTM model, which has been developed as a result of a Task Force on Toxic Impacts under the UNEP-SETAC Life Cycle Initiative, is a powerful tool for calculating CFs. It is a way to characterize human and ecotoxicological impacts in LCIA and comparative risk assessment (CRA) analysis. USEtoxTM was designed to describe the fate, exposure and effects of chemicals (Huijbregts et al., 2010a). USEtoxTM provides a parsimonious and transparent tool for human health and ecosystem CFs estimation. Based on a referenced database, it can be used to calculate CFs for several thousands of substances and forms the basis of the recommendations from UNEP-SETAC’s Life Cycle Initiative regarding the characterization of toxic impacts in LCA (Rosenbaum et al., 2008). Despite the large number of substances that have been considered in the USEtoxTM database (more than 3000 in the USEtoxTM organic database 1.01), only a small amount of these compounds are PPCPs (approximately sixty compounds of the organic database correspond to PPCPs). Therefore, the CFs of many PPCPs have not yet been calculated. LCIA conducted in systems with PPCPs may be incomplete or unrealistic if these compounds are not considered. Therefore, the estimation of CFs is a very important issue and a novel contribution in this research field. For this reason, Alfonsín et al. (2014) provide CFs for the toxicity related impact categories in LCA for 23 PPCPs. Some of the CFs already available in databases were updated (11 compounds) whereas others were implemented for the first time by means of USEtoxTM (12 compounds) and USES-LCA 2.0 methodologies. They cited only five previous studies that calculated a limited number of PPCPs' CFs by different methodologies. More recently, Roos et al. (2017), provided a set of 72 CFs, calculated with USEtoxTM, for some of the most common textile chemicals, in order to include them in LCA studies for this type of industry. This indicates that there is still a lack of data on the CFs of many compounds, which, therefore, cannot be included in LCA studies. Hence, in this paper, the CFs for 27 PPCPs have been calculated following the USEtoxTM methodology, to complement its own database and thus be able to incorporate these compounds into LCIA studies. Six compartments were considered as to which type of emission can take place (continental urban air, continental rural air, continental freshwater, coastal seawater at continental scale, continental natural soil and continental agricultural soil) and toxicity potentials were estimated for two different impact categories: human health and freshwater aquatic environments, according to the scope of USEtoxTM design. Additionally, using the new CFs calculated in this work, those CFs existing in the software database of USEtoxTM, the new CFs calculated by Alfonsín et al. (2014) and the data of occurrence (emissions) of PPCPs in the aquatic environments in Spain (Ortiz et al., 2013a), the human toxicity and ecotoxicity impact scores (IS) for 49 PPCPs have been estimated. These ISs have been used to develop and analyze ranking scores from CFs and then, compare the relative toxicity between these compounds and compare with other ranking of concern such as the established in Ortiz et al. (2013b). 2. Materials and methods The main steps carried out for the realization of this work are summarized below. The procedure for collecting the input parameters, the equations used, the run of the software and the results analysis of the USEtoxTM model were carried out according to the published literature (Huijbregts et al., 2005a; 2005b; 2010a; 2010b; Rosenbaum et al., 2008) and are shown in Figure 1 Figure 1. Flow diagram of main steps for the calculation of CFs. Adapted from Huijbregts et al. (2010) USEtoxTM considers environmental compartments to be well-mixed boxes that contain and exchange contaminant mass. Their total mass, total volume, solid-phase mass, liquid-phase mass and gas-phase mass, describe the compartment. Contaminants move between, and are transformed within compartments through a series of transport and transformation processes that can be represented mathematically by first-order processes, which depend on the physicochemical characteristics of the chemicals modelled. In each compartment, different factors are considered in the USEtoxTM model: persistence, transportation between compartments by cross-media transfers (dispersive, advective as volatilization, precipitation, etc.), transformation by a physical, chemical, or biological degradation process, or the irreversible removal from a compartment by leaching and/or burial (Fantke et al. 2017). All these considerations, and the values and parameters that emerge from them, define the final results of CFs and the differences observed between compartments. It is necessary to highlight that the USEtoxTM program has limitations that should be considered when it is used. The primary outputs of USEtoxTM are characterization factors for human toxicity and freshwater ecotoxicity; other impact categories are not considered. This assumes that the compartments are homogeneous. It does not account for speciation or other potentially important specific processes for metals, metal compounds and certain types of organic chemicals. It does not allow for the degradation of vegetation in the exposure model, and there are uncertainties regarding input data (Fantke et al. 2017). 2.1. Selection of PPCPs Similar to our previous pieces of research (Ortiz et al. 2013a, 2013b; 2014), the PPCPs selected are some of the most worldwidely important pharmaceutical active compounds (PhACs) and PCPs. Their consumption and occurrence in aquatic environments and in wastewater treatment plants (WWTPs) is relevant, and there is evidence of their potential ecotoxicity in the different compartments of the environment. Ortiz et al. (2013a) found that acetaminophen, amoxicillin, valsartan, omeprazole, clarithromycin, galaxolide, tonalide, iopromide and iohexol (among other compounds) had the highest level of occurrence in the Spanish aquatic environment. Additionally, in Ortiz et al. (2013b) a persistence, bioaccumulation and toxicity ranking score was carried out, and some of the compounds found at the top were: galaxolide, tamoxifen, sertraline, atorvastatin, tonalide, triclosan, irbesartan, fluoxetine, paroxetine and 17-α ethynylestradiol. When occurrence level was considered in the ranking score, tamoxifen was highlighted. In general, Ortiz et al. (2013b) found that fragrances, hormones, antidepressants, anxiolytics and blood lipid regulators presented the highest levels of risk. On the other hand, an environmental risk assessment of 26 PPCPs were done by Ortiz et al. (2014) with both new and current ecotoxicity values, and at least half of the compounds being studied were cataloged as being harmful to aquatic organisms, among them: triclosan, omeprazole, methylparaben, ethylparaben, propylparaben, clofibrate, ciprofloxacin, clarithromycin, diclofenac, naproxen and norfloxacin. Nevertheless, it is necessary to emphasize that some CFs of the PPCPs are already in the USEtoxTM database and are not included in this study because there is no new data (according to our knowledge) that would make it possible to update these values. Therefore, twenty seven PPCPs were chosen in order to have their first, new CFs calculated, for subsequent use in LCA studies. These compounds can be seen in the first row of Table 1. With the purpose of fulfilling the second goal of this study (to estimate the human toxicity and ecotoxicity ISs in the Spanish case study), twenty two values of CFs (calculated with the same tool, USEtoxTM) were taken from the bibliography (Alfonsin et al., 2014 and the database of USEtoxTM) in order to make a comparative study of ISs with as many compounds as possible (second and third row in Table 1). Therefore, a total of 49 compounds from 14 different therapeutic classes have been considered in this study: analgesic/antipyretic (1), angiotensin converting enzyme inhibitor (1), angiotensin receptor blockers (2), antibiotics (11), antidepressants (3), antiepileptics (4), anxiolytics (3), blood lipid regulators (3), cytostatics/cancer therapeutic (2), H2 blocker (1), hormones (4), platelet inhibitor (1), non-steroidal antiinflammatory drugs (NSAIDs)/antirreumatics (4), X-ray contrast media (3) and PCPs (6). Table 1. Pharmaceutical and personal care products under study of data for inhalation as a route of exposure); thus, the CFs of human toxicity calculated in this research should be taken as “interim”. Equations 11 to 16 (Table 2) are used to calculate the human-toxicological EFs. The summary of the calculation is as follow: results of chronic exposure in different animals were converted to ED50 (mg·kg-1·d-1) and then it was transformed into kg·person-1·lifetime-1 with equations 12, 13, 14, 15 and 16 (Table 2). For each PPCP, the geometric mean of these values was calculated and finally, the effect factor was determined with equation 11 (Table 2). 2.2.4. Input data Table 3 shows the input parameters that must be supplied by the user to the USEtoxTM program in order to estimate the CFs of the 27 PPCPs. These parameters are molecular weight (MW), partition coefficient between octanol and water (Kow), partition coefficient between organic carbon and water (Koc), Henry law coefficient at 25ºC (KH), vapor pressure at 25ºC (Pvap), solubility at 25ºC (Sol), degradation rate in air (KdegA), degradation rate in water (KdegW), bioaccumulation factor of the chemical (BAF), water ecotoxicity (as HC50) and human carcinogenic and non-carcinogenic effects of compounds under study (ED50ing, NonCancer and ED50ing, Cancer). In this study, whenever possible, the experimental values of physicochemical properties (MW, Kow, Koc, KH, Pvap and Sol) and BAF have been taken. However, when this was not possible, estimated values from EPI SuiteTM were used as recommended by Huibregts et al. (2010a, 2010b). For air degradation rates, experimental values for the hydroxyl radical rate constant (KOH) are available for some chemicals in EPI Suite™. To derive the KdegA, the KOH was multiplied by the hydroxyl radical concentration [OH•]. The default [OH•] was set at 1.5∙106 molecules (radicals)·cm-3 per 12h of daylight (US EPA, 2012). KdegW, soil (KdegSl) and sediment (KdegSd) were estimated by biodegradation half-life with EPI Suite™, and the Biowin3 model is used for USEtoxTM input to convert the ultimate biodegradation probability into half-lives for all chemicals in the database. In addition, division factors of 1:2:9 are used to extrapolate biodegradation rates for water, soil and sediment compartments respectively, as is suggested in EPI Suite™ (Huijbregts et al., 2010b). Water ecotoxicities for different species and trophic levels (bacteria, algae, crustacean, rotifer, mollusc and fish) were collected, as was explained in the effect factor section. The calculation steps of the logHC50 (α) according to Huijbregts et al. (2010a) are: (i) gather experimental or estimated EC50 data for the chemical of interest; (ii) specify for every EC50-value whether it is chronic or acute exposure; (iii) calculate the geometric mean chronic or acute EC50 for every individual species (in case of acute EC50-data, derive the chronic-equivalent EC50 per species by dividing by a factor of 2, acute-to-chronic extrapolation factor) and (iv) take the log of the geometric mean EC50s and calculate the average of the log-values (this average equals the logHC50). Non-carcinogenic data for rat, mouse, rabbit, dog and monkey were collected from different sources (World Health Organization database, European Agency for the Evaluation of Medicinal Products (EMEA) reports, toxicology studies of the US Food and Drugs Administration (FDA), reports of Scientific Committee on Consumer Safety of European Commission, United States Pharmacopeia monographs, material safety data sheet of Merck, Pifzer, Medsafe and La Roche and Rashmi et al., 2012). From these sources, the daily dose that causes a disease probability of 50% of population (ED50) was estimated from NOAEL or the LOAEL (chronic toxicity data). Carcinogenic data (as ED50) was obtained from Brambilla et al. (2012). In this study, the steps undertaken for the ED50 estimation according to Huijbregts et al. (2010a) guidelines were: (i) gather experimental non-carcinogenic oral ED50 data, (ii) specify for every ED50-value whether it is chronic, subchronic or subacute exposure; (iii) as the chronic value is needed, subchronic and subacute data have to be extrapolated to chronic ED50. According to the type of ED50-data found in the literature or database (non-human ED50-data, NOAEL or LOAEL), the chronic-equivalent ED50 must be derived using equations 12 to 16 (Table 2). The USEtoxTM software calculates the remainder of the necessary data from the input parameters referred to above, supplied by the user. More information about procedure, equations, considerations, estimations and the methodology can be consulted in the main literature that supports this software (Huijbregts et al., 2005a; 2005b; 2010a; 2010b; Rosenbaum et al., 2008). Table 3. Input parameters required for the ecotoxicological and human toxicity characterization factors calculation in USEtoxTM for the pharmaceuticals and personal care products under study Compound CAS Physicochemical parameters Degradation rates Ecotoxicity Human Toxicity● Bioaccumulatio n MW* Kow** Koc*** KH+ Pvap++ Sol.+++ KdegA† KdegW†† KdegSd††† KdegSl†††† α‡ ED50ing, NonCancer●● ED50ing, Cancer●● BAF●●● g/mol L kg-1 Pa m3 mol- 1 Pa mg L-1 s-1 s-1 s-1 s-1 log (mg L- 1) kg/lifetime/person kg/lifetime/perso n L kg-1 Acetaminophe n 103-90-2 151.1 7 2.88 45.09 6.51∙10-8 2.59∙10-4 14000 2.65∙10- 5 5.30∙10- 7 5.89∙10- 8 2.65∙10- 7 1.8800 673 NC 0.98 Alprazolam 28981-97-7 308.7 7 131.82 98320 5.19∙10-5 2.21∙10-6 13.10 1.14∙10- 5 2.10∙10- 7 2.33∙10- 8 1.05∙10- 7 -0.4360 8.11 NC 14.10 Amoxicillin 26787-78-0 365.4 1 7.41 108.40 2.52∙10-16 6.26∙10- 15 4000.0 0 2.08∙10- 4 2.10∙10- 7 2.33∙10- 8 1.05∙10- 7 1.7800 1960 NC 1.10 Atorvastatin 134523-00- 5 558.6 6 229086 7 28570 4.64∙10-17 9.26∙10- 23 0.0011 3.42∙10- 4 1.30∙10- 7 1.44∙10- 8 6.50∙10- 8 -0.6160 30.70 1.23 104.00 Azithromycin 83905-01-5 749.0 0 10471 3135.00 4.27∙10-18 3.53∙10- 22 0.0620 6.35∙10- 4 4.50∙10- 8 5.00∙10- 9 2.25∙10- 8 0.0669 16.20 NC 12.50 Bromazepam 1812-30-2 316.1 6 112.20 3605.00 4.57∙10-7 2.53∙10-7 175.20 8.68∙10- 6 1.30∙10- 7 1.44∙10- 8 6.50∙10- 8 0.7080 103.00 NA 9.94 Cefaclor 53994-73-3 367.8 1 2.24 104.30 9.15∙10-13 2.96∙10- 13 10000 2.02∙10- 4 2.10∙10- 7 2.33∙10- 8 1.05∙10- 7 3.0600 2680.00 NC 0.98 Ciprofloxacin 85721-33-1 331.3 5 1.91 10.00 1.10∙10-12 3.80∙10- 11 30000 4.70∙10- 4 1.30∙10- 7 1.44∙10- 8 6.50∙10- 8 0.4450 506.00 NA 0.98 Clarithormycin 81103-11-9 747.9 7 1445.44 149.40 6.77∙10-20 3.10∙10- 23 0.3420 5.97∙10- 4 4.50∙10- 8 5.00∙10- 9 2.25∙10- 8 -0.1480 344.00 NA 15.30 Enalapril 75847-73-3 376.4 6 1.17 348.50 2.92∙10-10 1.41∙10- 10 16400 1.77∙10- 4 5.30∙10- 7 5.89∙10- 8 2.65∙10- 7 2.0000 353.00 NC 0.916 Ethylparaben 120-47-8 166.1 8 295.12 162.18 4.86∙10-4 1.24∙10-2 885.00 1.89∙10- 5 5.30∙10- 7 5.89∙10- 8 2.65∙10- 7 0.8930 3930.00 NC 8.15 Gabapentin 60142-96-3 171.2 4 0.0794 53.14 9.42∙10-7 3.91∙10-8 4491.0 0 6.02∙10- 5 5.30∙10- 7 5.89∙10- 8 2.65∙10- 7 3.5100 981.00 436.22 0.90 Iohexol 66108-95-0 821.1 5 0.0009 10.00 4.17∙10-26 5.41∙10- 27 106.50 1.04∙10- 4 1.30∙10- 7 1.44∙10- 8 6.50∙10- 8 4.1000 29.70 NA 0.89 Iopamidol 60166-93-0 777.0 9 0.0038 10.00 2.26∙10-27 1.78∙10- 28 140000 8.66∙10- 5 1.30∙10- 7 1.44∙10- 8 6.50∙10- 8 3.1800 68.90 NA 0.89 Irbesartan 138402-11- 6 428.5 4 204173 133700 0 1.17∙10-9 1.64∙10- 13 0.0599 5.58∙10- 5 2.10∙10- 7 2.33∙10- 8 1.05∙10- 7 -1.3100 43.30 NC 2480 Ketorolac 74103-06-3 255.2 8 208.93 428.80 8.74∙10-6 1.96∙10-5 572.30 3.05∙10- 4 5.30∙10- 7 5.89∙10- 8 2.65∙10- 7 0.6960 12.50 NC 22.00 Levofloxacin 100986-85- 4 361.3 8 0.4074 0.99 1.68∙10-12 1.31∙10- 10 28260 2.95∙10- 4 4.50∙10- 8 5.00∙10- 9 2.25∙10- 8 0.9650 38.20 NC 0.90 Lorazepam 846-49-1 321.1 6 245.47 944.40 1.58∙10-9 4.11∙10- 10 80.00 1.68∙10- 5 1.30∙10- 7 1.44∙10- 8 6.50∙10- 8 0.8950 67.10 NC 25.00 Methylparaben 99-76-3 152.1 5 91.20 86.29 2.90∙10-3 1.14∙10-1 2500.0 0 1.66∙10- 5 5.30∙10- 7 5.89∙10- 8 2.65∙10- 7 1.0600 4480.00 NC 3.88 Norfloxacin 70458-96-7 319.3 4 0.0933 18.68 1.99∙10-12 1.11∙10-9 177900 4.82∙10- 4 1.30∙10- 7 1.44∙10- 8 6.50∙10- 8 1.0200 188.00 NA 0.890 Omeprazole 073590-58- 6 345.4 2 169.82 1455.00 3.08∙10-14 1.55∙10-9 82.28 1.43∙10- 4 1.37∙10- 8 1.52∙10- 9 6.85∙10- 9 -0.1690 21.70 3.49 3.46 Paroxetine 61869-08-7 329.3 7 8912.50 12360 5.96∙10-5 6.39∙10-6 35.27 2.45∙10- 4 1.30∙10- 7 1.44∙10- 8 6.50∙10- 8 -0.4580 3.93 4.36 624.00 Pregabalin 148553-50- 8 159.2 3 0.0166 25.05 2.19∙10-9 2.69∙10-7 19630 6.12∙10- 5 5.30∙10- 7 5.89∙10- 8 2.65∙10- 7 3.7500 139.00 NC 0.89 Propylparaben 94-13-3 180.2 1 1096.48 286.60 1.39∙10-2 4.09∙10-2 500.00 2.11∙10- 5 5.30∙10- 7 5.89∙10- 8 2.65∙10- 7 0.4330 39.30 NC 15.60 Sertraline 79617-96-2 306.2 4 194984 170800 1.36∙10-2 1.56∙10-4 3.52 1.47∙10- 4 1.30∙10- 7 1.44∙10- 8 6.50∙10- 8 -0.4870 34.20 1.23 50800 Simvastatin 79902-63-9 418.5 8 47863 10940 3.09∙10-7 5.65∙10- 10 0.0300 3.44∙10- 4 2.10∙10- 7 2.33∙10- 8 1.05∙10- 7 -0.3970 109.00 5.45 151.00 Valsartan 137862-53- 4 435.5 3 4466.84 22630 3.38∙10-11 1.09∙10- 13 1.41 6.42∙10- 5 5.30∙10- 7 5.89∙10- 8 2.65∙10- 7 0.2420 3.93 NA 215.00 *Molecular weight. **Partitioning coefficient between octanol and water. ***Partitioning coefficient between organic carbon and water. + Henry law coefficient at 25ºC. ++ Vapor pressure at 25ºC. +++ Solubility in water at 25ºC. † Degradation rate in air. †† Degradation rate in water. Results of BIOWIN3  Biodegradation rates in USEtoxTM: Hours 4.7∙10-5; hours to days  6.4∙10-6; days 3.4∙10-6; days to weeks  9.3∙10-7; weeks 5.3∙10-7; weeks to months 2.1∙10-7; months 1.3∙10-7; recalcitrant 4.5∙10-8. ††† Degradation rate in sediment. †††† Degradation rate in soil. ‡ log of HC50 (HC50: Hazardous concentration of chemical at which 50% of the species are exposed above their EC50. The EC50 is the water PPCP concentration at which 50% of a population displays an effect). ●According to the USEtoxTM methodology the human toxicity is calculated for carcinogenic and non-carcinogenic effects and for inhalation and ingestion (exposure routes). This table only shows human toxicity for non-carcinogenic effect by ingestion, due to the lack of data of cancer effect and inhalation route. For more information see methodology and discussion section. ●● Daily dose (by ingestion) of PPCP that causes a disease (carcinogenic or non-carcinogenic) probability of 50% in a person in its lifetime. ●●●Bioaccumulation factor of the chemical in fish estimated by EPI SuiteTM (US EPA, 2012) . NC No carcinogenic . NA: Not available 2.3. Spanish toxicity impact scores (IS) In LCA, a toxicity impact score (IS) is calculated with equations 1 and 6 as reported in Table 2. This value is the summation of the product of mass emitted per emission scenario, by the corresponding toxicity CF, taking into account all the scenarios where the pollutant is emitted (Huijbregts et al. 2010a; Rosebaum et al. 2008). IS allows us to group into a single index the impact (ecotoxicity or human toxicity) of a compound released into the different compartments of nature. The equation and procedure for calculating IS for ecotoxicity (Eq. 1) and for human toxicity (Eq. 6) are summarized in Table 2. The ISs are reported as comparative toxic units (CTU) that can be compared with ISs obtained from other methodologies. In a recent study, Ortiz et al. (2013a) estimated the occurrence (mass·year-1) of 88 PPCPs and metabolites in the aquatic environment in Spain. These results have been used to estimate ecotoxicological and human toxicity ISs (in CTU·year-1) of the PPCPs considered. A mass balance approach was used by Ortiz et al. (2013a) in order to estimate their occurrence in aquatic environments, and the data for their removal in WWTPs was assessed by STPWIN™, a special module of EPI Suite™. STPWIN™ predicts the removal of a chemical in a typical activated sludge-based sewage treatment plant. Values are given for total removal and three processes that may contribute to removal: biodegradation, sorption to sludge, and air stripping (US EPA, 2012). The program assumes a standard system design and set of default operating conditions, and takes physico-chemical parameters from EPI SuiteTM, which works with a large database of experimental values, or estimates them with quantitative structure-activity relationship (QSAR) models. EPI SuiteTM has facilitated the calculation of the mass of PPCPs adsorbed in the sludge and volatilized to the air. In this way, a total ecotoxicological IS has been obtained, that includes three compartments: water, soil and air. For this case study, emissions were considered in the following environmental compartments: continental freshwater (water), continental natural soil (soil) and continental urban air (air). 3. Results and discussion 3.1. Discussion of input data This section briefly shows the analysis of the values of the input parameters taken or estimated in order to obtain human and ecotoxicological CFs for the PPCPs under study. 3.1.1. Physico-chemical parameters In general, the PPCPs under study present variable degrees of solubility, Koc and Kow. These parameters provide an estimation of the mobility of the PPCPs in water environments, soils and sediments. Atorvastatin, azithromycin, clarithromycin, irbesartan, paroxetine, simvastatin, valsartan and sertraline show the highest values of Kow and Koc, and low solubility; therefore, these compounds will be probably located in soils or sediments, or bioaccumulated. Degradation rates in water, soils and sediments are of the same order of magnitude, although these are slightly higher in water than in soil and sediments, in most cases. Degradation rates in air are the highest among all the compartments, possibly due to the different photochemical effects and reactions that take place in this compartment. Despite this, all compounds present low Pvap (lower than 1 Pa) and KH, which indicates that they will not be found in significant quantities in the air. The compounds with the highest values of bioaccumulation (estimated by EPI SuiteTM, US EPA, 2012) were: sertraline, irbesartan, paroxetine, valsartan and simvastatin. 3.1.2. Ecotoxicological effects Usually, fish, crustaceans and algae are the main representative organisms that are considered when estimating the logHC50 values (α) (Table 3), but in the present study, other organisms (molluscs, bacteria and rotifers) were incorporated where data was available. The logHC50 of alprazolam, azithromycin, bromazepam, gabapentin, iohexol, iopamidol, irbesartan, ketorolac, lorazepam, pregabalin, simvastatin and valsartan were estimated only with fish, crustaceans and algae presenting acute and chronic toxicity. Sertraline, omeprazole, amoxicillin, cefaclor and levofloxacin logHC50 were calculated from four species (the main three organisms and bacteria with acute toxicities), while acetaminophen and clarithromycin presented toxicities in fish, algae, crustaceans, rotifers and bacteria. Values for the acute and chronic toxicities caused by parabens in crustaceans, fish and bacteria were used to estimate its logHC50. Norfloxacin toxicities present in fish, algae, rotifers and bacteria were used for estimating its logHC50. The remaining four compounds (paroxetine, enalapril, ciprofloxacin and atorvastatin) only had two values of toxicity (algae and crustacean). Several levels of toxicity were found: Median lethal dose concentration; Half maximal effective concentration; Half maximal inhibitory concentration; No observed effect concentration; and Lowest observed effect concentration (for more information see complementary material). However, the PPCPs considered in this study did not have the same quantity of ecotoxicity data; so it would be important to complement this information for compounds that present high toxicity. Iohexol, pregabalin, gabapentin, iopamidol, cefaclor and enalapril were the pharmaceuticals with the highest values of logHC50. 3.1.3. Human Health effects It is known that chemicals may pose hazards to organisms including humans, as indicated by observable effects (e.g. in vivo and in vitro bioassays). Antibiotics is one of these cases; although the ideal antibiotic is toxic to bacteria without affecting humans/animals, the reality is more complicated, and directly toxic side effects are common for several classes of antibiotics at doses used for therapy. A few, relatively persistent antibiotics have been found in drinking-water at very low ng·L-1 levels. The greatest concern about antibiotics in the environment is their potential role in promoting resistance development in human and animal pathogens (Larsson, 2014a). Different available databases include human health effects that are generally an approximation of bioassays in some typical species used for this purpose. In addition, there is currently a wide range of endpoints available from predictive quantitative structure–activity relationship (QSAR) models driven by many different computational software programs and data sources grouped under the term “in silico toxicology” (Valerio, 2009). These tools also are used for PPCPs that are already on the market and for estimating their human effects. In USEtoxTM, and therefore in this research, human toxicity includes carcinogenic and noncarcinogenic effects. Alprazolam, atorvastatin, azithromycin, clarithromycin, irbesartan, omeprazole, paroxetine, sertraline and simvastatin are the compounds with the highest human toxicity according to the USEtoxTM methodology. 3.1.3.1. Carcinogenic effects In this study, as it can be seen in Table 3, six compounds present evidence of carcinogenic effects (atorvastatin, gabapentin, omeprazole, paroxetine, sertraline and simvastatin) in rats or mice during long term studies (chronic), fourteen have no evidence of carcinogenic effects (acetaminophen, alprazolam, amoxicillin, azithromycin, cefaclor, enalapril, ethylparaben, irbesartan, ketorolac, levofloxacin, lorazepam, methylparaben, pregabalin and propylparaben), and for seven compounds there was no available data (bromazepam, ciprofloxacin, clarithromycin, iohexol, iopamidol, paroxetine and valsartan) according to the information (ED50) reported in Brambilla et al. (2012) and the database consulted and recommended by the USEtoxTM users’ manual. The assessment of the carcinogenic potential of pharmaceuticals and the evaluation of their potential risk to humans is a major challenge for the scientific community, industry and regulatory agencies. The importance of reaching appropriate conclusions and balancing those conclusions with benefits, and the potential impact that those decisions may have on public health cannot be overstated (DeGeorge, 1998). Abraham and Ballinger (2012) affirm that human exposure to pharmaceuticals can cause cancer, so modern societies have assessed the carcinogenicity of new drugs since the 1960s. Recent studies provide evidence of the carcinogenic effect of some PPCPs including estrogens, analgesic mixtures with phenacetin and antineoplastic drugs (Grosse et a l. 2009). Brambilla and Martelli (2009) made a compendium of the genotoxic and carcinogenic information of 838 marketed drugs, whose expected clinical use is for a continuous period of at least 6 months, or intermittent over an extended period of time. Of these 838 drugs, 472 (56.3%) have at least one positive test result for genotoxicity or carcinogenicity, a fairly high percentage for this type of chemical compounds. These studies serve as an experimental basis for asserting that PPCPs can cause carcinogenic effects, so therefore this information should be considered and included in the database of USEtoxTM and subsequently used in LCA studies. The traditional approach to testing the carcinogenicity of pharmaceuticals is relatively standardized. It relies on testing the maximum tolerated dose (MTD) on usually two rodent species for 2 years. The results of these studies were viewed as either positive or negative, with only minimal attempts to evaluate the relevance of the findings for humans (DeGeorge, 1998). In this sense, Abraham and Ballinger (2012) worked on the validation and application of new techno-regulatory testing standards, specifically using genetically-engineered mouse (GEM) models in pharmaceutical carcinogenic risk assessment. This methodology or other more traditional ones, experimental or not, may be used in order to find out or predict the possible carcinogenic potential of PPCPs and to thus include carcinogenic data in studies of environmental risk/hazard assessment such as LCIA. Data for human cancer has been found in lower quantities than for other effects; and it only remains for atorvastatin, gabapentin, omeprazole, paroxetine, sertraline, and simvastatin to have their value for ED50ing, Cancer calculated. Despite this, there are not enough studies that provide evidence as to whether many PPCPs are carcinogenic or not, and that list the minimum doses that cause this adverse effect. In this research, there was a lack of data regarding carcinogenic effects for 30% of the PPCPs under study. 3.1.3.2. Non-carcinogenic effects PPCPs must undergo strict controls before approving their use on animals and humans. Noncarcinogenic effects on humans or non-humans are some of those important pieces of data that must be reported for the safe use of these compounds. In this study, for all the PPCPs, non-carcinogenic data was available as NOAEL or LOAEL, and was reported for various species (mouse, rabbit, rat, dog and monkey) and for different lengths of exposure time (sub-acute, sub-chronic and chronic). These data were converted to the chronicequivalent ED50 by using Eqs. 15 and 16 (Table 2). The ingestion route was only considered due to the lack of data for other routes of exposition, such as inhalation. Affectation of the liver, kidney, testicle, lung, eyes and central nervous system, and symptoms such as sedation, ataxia, convulsive seizures, abnormal secretion of sex hormones, decrease in blood pressure, hyperplasia of the juxtaglomerular apparatus, pallor, hematologic and pathologic alterations, benign tumors, cardiovascular malformations, embryotoxicity and teratogenicity were the main adverse effects reported in the literature consulted, and cited in the material and methods section for non-carcinogenic effects. The effects shown in Table 4 relate to information obtained mainly from the material safety data of each compound, and from reports by the World Health Organization, EMEA and FDA. Table 4. PPCPs non-carcinogenic effects on different species* Compound Non-carcinogenic effects Specie Acetaminophen Sub chronic effects Mice (males and females) and rat Alprazolam Embryotoxicity and teratogenic effects Rabbit and rat Amoxicillin and bromazepam Reproductive effects Rat Atorvastatin Effects on the liver Dog, rat and mouse Azithromycin Dog and rat Clarithromycin and bromazepam Dog Gabapentin and sertraline Rat Bromazepam Sedation, ataxia, convulsive seizures Dog Clarithromycin Cardiovascular malformation Rat Ethylparaben Secretion of sex hormones Rat Irbesartan Decreases in blood pressure and hyperplasia of the juxtaglomerular apparatus Rat and monkey Ketorolac Hematologic and pathologic effects Monkey Levofloxacin Impaired pup survival and decreased birth weight Rat Lorazepam Fetotoxicity Rabbit Pregabalin Adverse effects on blood forming organs and peri/postnatal developmental Rat Sertraline Benign tumors Mouse Effects on central nervous system Dog Early embryonic development and developmental toxicity Rat Simvastatin Eyes Rat Eyes and central nervous system Dog Valsartan Renal negative effects Rat *All these effects were reported by oral route of administration except for lorazepam that was intravenous Methylparaben, ethylparaben, cefaclor, amoxicillin and gabapentin were the PPCPs with the highest values of non-cancerous human toxicity (ED50ing) parameters. They found a high mean hazard quotient (HQ) for all three-model species combined for cardiovascular, sedatives, hormones and gastrointestinal PPCPs. Despite the large difference in the amount of compounds evaluated in their study comparing to the present research, similarly, gastrointestinal drugs (e.g. omeprazole) and hormones showed a high toxicity impact score. Muñoz et al. (2008) estimated characterization factors for 98 frequently detected pollutants (approximately half of them were PPCPs), using two characterization models, EDIP97 and USES-LCA, and developed a LCIA-based ranking of the potential impacts of priority and emerging pollutants in urban wastewater. They found that PPCPs were very important contributors to the toxicity in WWTPs, with ciprofloxacin, fluoxetine and nicotine (not considered in our study) being the main PPCPs of concern. In the present study, these two first compounds also appear as sixth and thirteenth in the ranking of ecotoxicological potential impact calculated from USEtoxTM CFs. Hence, the IS of PPCPs highlights those compounds that may be of special interest in environmental impact studies such as LCA. It is important to highlight that, generally, all these methodologies and their rankings are based on the individual effect of each compounds, and the effect of the mixture has rarely been compared. In LCA studies, the contribution of each compound to each impact category is considered (e.g. global warming, ozone depletion, toxicity, acidification, energy and water use, etc.), but mixtures and complex interactions in the environment are not well characterized, especially in human and ecosystems. Toxicologists consider that compounds which affect the same organ can cause an additive effect; LCA do not work with target organisms, their impact categories are much broader, so it is more difficult for this methodology to manage the possible synergistic or antagonistic effects of mixtures (UNEP, 2003). In this sense, some researches were conducted in order to obtain CFs for evaluating the potential impacts of complex mixtures on ecosystems, such as the studies carried out by Bamard et al. (2011) who developed a method to calculated CFs for hydrocarbon mixtures, and Li et al. (2015), who improved health impact estimates of 16 polyclyclic aromatic hydrocarbons (PAH) with the USEtoxTM method. They explored the importance of emission profiles for the PAH mixture and illustrated how these improvements affect the LCIA case study; but for the majority of existing compounds which can reach the environment, be mixed and interact with it, much still remains to be studied. Table 7. Spanish toxicity impact score of PPCPs based on human health and ecotoxicity characterization factors. Compound Human health characterization factor (CTU*h·kg-1) and mass emitted in the different compartments (kg·year-1) Human toxicity impact score (CTU h + ·year-1) Ecotoxicity characterization factor (CTU*e·kg-1) and mass emitted in the different compartments (kg·year-1) Ecotoxicity impact score (CTU e + ·year- 1) ECUaira MAir** ECFWb Mwater** ECNSc MSoil** ECUaira MAir** ECFWb Mwater** ECNSc MSoil** 17α- Ethinylestradiol 6.79∙10-2 na 2.45∙10-2 0.29 3.02∙10-6 0.66 7.11∙10-3 3.02∙104 na 1.69∙106 0.29 2.56∙105 0.66 6.59∙105 17β-estradiol 7.76∙10-4 na 2.18∙10-3 69.12 3.02∙10-6 54.86 1.51∙10-1 3.30∙106 na 1.84∙108 69.12 2.56 54.86 1.27∙1010 Acetaminophen 5.6∙10-9 na 2.90∙10-8 23267 3.40∙10-9 453155 2.22∙10-3 5.07 na 1.25∙102 23267 1.47∙101 453155 9.57∙106 Alprazolam 1.39∙10-6 na 2.12∙10-6 60.21 6.31∙10-10 1.69 1.28∙10-4 5.39∙102 na 2.01∙104 60.21 5.99 1.69 1.21∙106 Amoxicillin 1.75∙10-8 na 2.11∙10-8 15257 2.74∙10-9 101325 5.99∙10-4 4.37∙101 na 3.33∙102 15257 4.32∙101 101325 9.45∙106 Atorvastatin 5.10∙10-5 na 4.72∙10-5 715.39 6.10∙10-8 1145.81 3.38∙10-2 1.76∙103 na 4.53∙104 715.39 5.85∙101 1145.81 3.24∙107 Azithromycin 4.60∙10-6 na 6.11∙10-6 1933.32 1.76∙10-7 958.03 1.20∙10-2 2.16∙103 na 3.68∙104 1933.32 1.06∙103 958.03 7.22∙107 Bromazepam 1.99∙10-7 na 5.60∙10-7 100.13 5.08∙10-9 2.72 5.61∙10-5 1.70∙102 na 5.00∙103 100.13 4.53∙101 2.72 5.01∙105 Carbamazepine 2.08∙10-6 na 7.68∙10-6 2595.31 1.12∙10-7 1204.28 2.01∙10-2 1.65∙101 na 8.54∙102 2595.31 1.25∙101 1204.28 2.23∙106 Cefaclor 1.24∙10-8 na 1.54∙10-8 119.60 2.06∙10-9 2.62 1.85∙10-6 2.27 na 1.71∙101 119.60 2.28 2.62 2.05∙103 Ciprofloxacin 1.11∙10-7 na 1.13∙10-7 2402.03 4.46∙10-8 12957 8.50∙10-4 3.05∙103 na 9.84∙103 2402.03 3.88∙103 12957 7.39∙107 Clarithormycin 2.92∙10-7 na 3.14∙10-7 5820.23 8.60∙10-8 1669.22 1.97∙10-3 1.52∙104 na 6.46∙104 5820.23 1.77∙104 1669.22 4.05∙108 Clofibrate 2.07∙10-7 1.11∙10-2 3.67∙10-7 2.45 2.58∙10-8 0.56 9.16∙10-7 na 1.11∙10-2 na 2.45 na 0.56 na Cyclophosphamide 2.45∙10-6 na 7.33∙10-6 9.78 1.03∙10-6 119.86 1.95∙10-4 na na na 9.78 na 119.86 na Diclofenac 3.08∙10-7 na 1.22∙10-6 3963.10 4.84∙10-8 6613.79 5.16∙10-3 5.03∙101 na 2.67∙103 3963.10 1.05∙102 6613.79 1.13∙107 Enalapril 6.58∙10-9 na 5.55∙10-8 725.20 1.20∙10-9 1188.09 4.17∙10-5 2.01 na 9.40∙101 725.20 2.04 1188.09 7.06∙104 Erythromycin na na na 910.75 na 116.94 na 3.22∙103 na 2.49∙104 910.75 3.15∙103 116.94 2.30∙107 Estrone 2.64∙10-4 na 3.17∙10-4 28.22 5.37∙10-7 153.00 9.03∙10-3 4.39∙101 na 2.14∙104 28.22 1.93∙101 153.00 6.07∙105 Ethylparaben 1.14∙10-9 na 5.33∙10-9 na 2.32∙10-10 na na 3.85∙101 na 1.21∙103 na 5.23∙101 na na Fluoxetine 2.49∙10-5 na 2.6∙10-5 324.51 na 125.92 8.44∙10-3 3.82∙102 na 4.64∙104 324.51 7.32∙101 125.92 1.51∙107 Gabapentin 5.86∙10-9 na 6.53∙10-8 1943.87 6.83∙10-9 47406 4.51∙10-4 7.94∙10-2 na 2.94 1943.87 3.08∙10-1 47406 2.03∙104 Galaxolide 6.95∙10-7 na 5.00∙10-7 69221 4.69∙10-9 102389 3.51∙10-2 2.19∙101 na 1.01∙104 69221 1.72∙101 102389 7.01∙108 Ibuprofen 4.16∙10-7 6.74 3.71∙10-7 4849.50 1.74∙10-8 87853 3.33∙10-3 3.25 6.74 2.09∙102 4849.50 3.65 87853 1.33∙106 Iohexol 1.93∙10-6 na 1.93∙10-6 5127.22 7.59∙10-7 4691.52 1.34∙10-2 7.00∙10-1 na 2.17 5127.22 8.54∙10-1 4691.52 1.51∙104 Iopamidol 8.32∙10-7 na 8.30∙10-7 11416 3.27∙10-7 1296.93 9.90∙10-3 5.90 na 1.82∙101 11416 7.17 1296.93 2.17∙105 Iopromide 2.29∙10-7 na 1.86∙10-7 14752 7.29∙10-8 6202.81 3.20∙10-3 5.57 na 1.74∙101 14752 6.82 6202.81 2.99∙105 Irbesartan 7.28∙10-7 na 9.36∙10-7 3810.87 9.90∙10-11 23076 3.57∙10-3 6.39∙102 na 1.68∙104 3810.87 1.78 23076 6.42∙107 Ketorolac 4.62∙10-8 na 1.87∙10-6 217.64 3.33∙10-8 6.94 4.06∙10-4 3.48∙101 na 1.89∙103 217.64 3.37∙101 6.94 4.12∙105 Levofloxacin 2.17∙10-6 na 2.51∙10-6 4041.49 1.20∙10-6 87.97 1.03∙10-2 1.67∙103 na 4.99∙103 4041.49 2.38∙103 87.97 2.04∙107 Lorazepam 5.42∙10-7 na 1.02∙10-6 304.99 3.30∙10-8 10.25 3.11∙10-4 2.12∙102 na 3.42∙103 304.99 1.11∙102 10.25 1.05∙106 Methylparaben 1.04∙10-9 na 4.51∙10-9 2148.67 3.36∙10-10 56.43 9.70∙10-6 3.24∙101 na 8.27∙102 2148.67 6.02∙101 56.43 1.78∙106 Naproxen 1.42∙10-7 na 2.95∙10-7 4196.75 6.61∙10-9 12592 1.32∙10-3 3.94 na 2.18∙102 4196.75 4.86 12592 9.76∙105 Norfloxacin 1.20∙10-7 na 3.05∙10-7 1118.69 1.09∙10-7 2334.02 5.94∙10-4 3.68∙102 na 2.64∙103 1118.69 9.42∙102 2334.02 5.15∙106 Omeprazole 3.61∙10-5 na 4.25∙10-5 12992 6.38∙10-6 1388.34 5.61∙10-1 1.29∙104 na 1.63∙101 12992 3.92∙102 1388.34 7.55∙105 Paroxetine 2.92∙10-6 na 1.46∙10-4 58.60 4.06∙10-7 22.66 8.54∙10-3 1.10∙103 na 6.25∙104 58.60 1.74∙102 22.66 3.66∙106 Pregabalin 1.35∙10-8 na 1.42∙10-7 4175.19 2.34∙10-8 90.88 5.95∙10-4 5.36∙10-2 na 1.67 4175.19 2.76∙10-1 90.88 7.00∙103 Propylparaben 1.03∙10-7 na 5.62∙10-7 688.81 1.58∙10-8 51.53 3.88∙10-4 8.11∙101 na 3.44∙103 688.81 8.88∙101 51.53 2.37∙106 Roxythromycin na na na 34.24 na 4.38 na 9.84∙101 na 2.18∙103 34.24 2.21∙101 4.38 7.47∙104 Salicylic acid na na na 859.07 na 7984.66 na 1.35∙101 na 1.61∙102 859.07 2.82∙101 7984.66 3.63∙105 Sertraline 6.70∙10-5 na 4.77∙10-3 488.95 1.62∙10-6 99.09 2.33 2.43∙102 na 1.80∙104 488.95 6.11 99.09 8.82∙106 Simvastatin 4.32∙10-6 na 7.55∙10-5 1267.82 3.01∙10-8 2647.25 1.97∙10-2 1.39∙103 na 4.08∙104 1267.82 7.92∙101 2647.25 5.20∙107 Sulphametoxazole 3.24∙10-8 na 1.58∙10-7 2084.07 1.03∙10-8 2315.58 3.53∙10-4 6.07∙101 na 2.99∙103 2084.07 1.95∙102 2315.58 6.68∙106 Tamoxifen na na na 9.78 na 119.86 na 2.82∙102 na 1.99∙104 9.78 3.08 119.86 1.95∙105 Testosterone na na na 0.14 na 0.02 na 2.37∙102 na 1.30∙104 0.14 1.17∙102 0.02 1.82∙103 Tonalide 1.04∙10-6 9.43∙101 2.77∙10-5 11075 1.82∙10-7 35161 3.13∙10-1 3.00∙101 9.43∙101 1.20∙104 11075 4.26∙101 35161 1.34∙108 Triclosan 1.11∙10-7 na 2.21∙10-7 na 5.01∙10-10 na na 2.58∙103 na 1.06∙105 na 1.61∙101 na na Trimethoprim 9.16∙10-8 na 5.66∙10-7 44.57 2.29∙10-8 5.69 2.54∙10-5 9.11 na 4.74∙102 44.57 1.92∙101 5.69 2.12∙104 Valproic acid na 8.52 na 229.80 na 5645.91 na 2.14 8.52 1.21∙102 229.80 1.53∙101 5645.91 1.14∙105 Valsartan 3.86∙10-6 na 1.18∙10-5 20351 4.62∙10-9 4810.05 2.40∙10-1 1.67∙102 na 4.37∙103 20351 1.71 4810.05 8.89∙107 *Comparative toxic units. **Ortiz et al. (2013a). aEmission into continental urban air. b Emission into continental freshwater. c Emission in to continental natural soil. Compounds of this study are shaded Figure 1. Spanish human toxicity impact score (IShum) for the selected PPCPs. 0 0,05 0,1 0,15 0,2 0,25 0,3 0,35 Tonalide Valsartan 17β-estradiol Sertraline Omeprazole Galaxolide Carbamazepine Iohexol Azithromycin Levofloxacin Iopamidol Estrone Fluoxetine 17α-ethinylestradiol Diclofenac Paroxetine Irbesartan Ibuprofen Iopromide Acetaminophen Clarithormycin Naproxen Atorvastatin Simvastatin Ciprofloxacin Amoxicillin Pregabalin Norfloxacin Ketorolac Propylparaben Sulphametoxazole Lorazepam Cyclophosphamide Gabapentin Alprazolam Bromazepam Enalapril Trimethoprim Methylparaben Cefaclor Clofibrate Human toxicity impact score (CTUh· year-1 ) Compound 0 0,0005 0,001 0,0015 0,002 0,0025 0,003 0,0035 Ibuprofen Iopromide Acetaminophen Clarithormycin Naproxen Atorvastatin Simvastatin Ciprofloxacin Amoxicillin Pregabalin Norfloxacin Ketorolac Propylparaben Sulphametoxazole Lorazepam Cyclophosphamide Gabapentin Alprazolam Bromazepam Enalapril Trimethoprim Methylparaben Cefaclor Clofibrate Figure 2. Spanish ecotoxicity impact score (ISeco) for the selected PPCPs 1,E+00 1,E+01 1,E+02 1,E+03 1,E+04 1,E+05 1,E+06 1,E+07 1,E+08 1,E+09 1,E+10 1,E+11 17β-estradiol Galaxolide Clarithormycin Tonalide Valsartan Ciprofloxacin Azithromycin Irbesartan Simvastatin Atorvastatin Erythromycin Levofloxacin Fluoxetine Diclofenac Acetaminophen Amoxicillin Sertraline Sulphametoxazole Norfloxacin Paroxetine Propylparaben Carbamazepine Methylparaben Ibuprofen Alprazolam Lorazepam Naproxen Omeprazole 17α-ethinylestradiol Estrone Bromazepam Ketorolac salicylic acid Iopromide Iopamidol Tamoxifen Valproic acid Roxythromycin Enalapril Trimethoprim Gabapentin Iohexol Pregabalin Cefaclor Testosterone Ecotoxicity impact score (CTU·year-1) Compound Conclusion PPCPs are a large group of compounds which are present in all compartments of nature. It is impossible to analyze all the interactions and effects of these compounds on the environment; therefore, the use of LCA studies and risk/hazard assessments are very useful tools for predicting their ecotoxicological and human impacts/effects. In order to implement these methodologies, an estimation of CFs is needed, using routines such as USEtoxTM. With this in mind, 27 CFs have been calculated for PPCPs widely used at present. A Spanish ranking toxicity impact score (IS) was done for ecotoxicological and human toxicity for 49 PPCPs as a case study, using these 27 new CFs found and 22 others existing in literature, in combination with data regarding the occurrence of these compounds in the Spanish environment. Angiotensin receptor blocker (valsartan, irbesartan), blood lipid regulators (simvastatin atorvastatin), H2 blocker (omeprazole) and antidepressant (sertraline) were the pharmaceuticals with the highest human health CFs in the different compartments of emission. Omeprazole, antibiotics (clarithromycin, ciprofloxacin) antidepressant (paroxetine, sertraline) and parabens had the highest ecotoxicological CFs. This study has established that emissions into continental freshwater originate the highest CFs, for both human and ecological impacts. Ecotoxicological CFs were much higher than human toxicity CFs, since the human tolerance of PPCPs is higher than for environmental biota. In the case of this study, the toxicity impact scores derived from the USEtoxTM CFs place fragrances, hormones, antibiotics, antidepressants, angiotensin receptor blockers and blood lipid regulators at the top of the ranking, similar to other rankings generated with other methodologies. The CFs of PPCPs estimated in this work offer the possibility of incorporation into new LCIA of LCA studies, or to formulate a ranking impact score list. Although this study focused on PPCPs, it is important to highlight the fact that other emerging micropollutants such as pesticides, alkylphenols, perflourinated compounds and various industrial organic chemicals are highly polluting, so it is recommended that CF estimation should continue for all these types of compounds, in order to thus include them in LCA studies. Acknowledgements The authors would like to thank the Regional Government of Castilla y León (Project VA067U16 and UIC71) for the support provided during this research and the Carabobo University, Venezuela for the PhD scholarship grants (No. CD-3417 and No. CD-2155). Conflict of interest The authors declare that they have no conflicts of interest. References Abraham J, Ballinger R (2012) Science, politics, and health in the brave new world of pharmaceutical carcinogenic risk assessment: Technical progress or cycle of regulatory capture?. Soc Sci & Med 75:1433-1440. doi: 10.1016/j.socscimed.2012.04.043 Alfonsín C, Hospido A, Omil F, Moreira MT, Feijoo G (2014) PPCPs in wastewater e Update and calculation of characterization factors for their inclusion in LCA studies. J Clean Prod 83:245-255. doi: 10.1016/j.jclepro.2014.07.024 Bamard E, Bulle C, Deschênes L (2011) Method development for aquatic ecotoxicological characterization factor calculation for hydrocarbon mixtures in life cycle assessment. Environ Toxicol Chem 30(10):2342-2352. doi: 10.1002/etc.632 Brambilla G, Martelli A (2009) Update on genotoxicity and carcinogenicity testing of 472 marketed pharmaceuticals. Mut Res 681:209-229. doi:10.1016/j.mrrev.2008.09.002 Brambilla G, Mattioli F, Robbiano L, Martelli A (2012) Update of carcinogenicity studies in animals and humans of 535 marketed pharmaceuticals. Mut Res 750(1): 1-51 Brausch JM, Rand GM (2011) A review of personal care products in the aquatic environment: Environmental concentrations and toxicity. Chemosphere 82:1518-1532. doi: 10.1016/j.chemosphere.2010.11.018 Cleuvers M (2003) Aquatic ecotoxicity of pharmaceuticals including the assessment of combination effects. Toxicol Let 142:185-194. doi:10.1016/S0378-4274(03)00068-7 Cleuvers M (2004) Mixture toxicity of the anti-inflammatory drugs diclofenac, ibuprofen, naproxen, and acetylsalicylic acid. Ecotox Environ Saf 59:309-315. doi:10.1016/S0147-6513(03)00141-6 Cooper ER, Siewicki TC, Phillips K (2008) Preliminary risk assessment database and risk ranking of pharmaceuticals in the environment. Sci Tot Environ 398:26-33. doi:10.1016/j.scitotenv.2008.02.061 Daughton CG, Brooks BW (2010) Active Pharmaceutical Ingredients and Aquatic Organisms. Available online at: http://www.epa.gov/esd/bios/daughton/APIs-aquaticbiota.pdf. Accessed on March 8, 2015. DeGeorge J (1998) Challenges in application of new approaches to carcinogenicity testing for pharmaceuticals. Toxicol Lett 102-103:565-568. PII S0378-4274(98)00249-5 Dobbins LL, Usenko S, Brain RA, Brooks BW (2009) Probabilistic ecological hazard assessment of parabens using Daphnia magna and Pimephales promelas. Environ Toxicol Chem 28(12):2744-2753. doi: 0730-7268/09 Ebele AJ, Abdallah MAE, Harrad S (2017) Pharmaceuticals and personal care products (PPCPs) in the freshwater aquatic environment. Emerging Contaminants 3:1-16. doi: 10.1016/j.emcon.2016.12.004 Fantke P (Ed.), Bijster M, Guignard C, Hauschild M, Huijbregts M, Jolliet O, Kounina A, Magaud V, Margni M, McKone TE, Posthuma L, Rosenbaum RK, van de Meent D, van Zelm R (2017) USEtox® 2.0 Documentation (Version 1), http://usetox.org Fent K, Weston AA, Caminada D (2006) Review Ecotoxicology of human pharmaceuticals. Aquat Toxicol 76:122–159. doi:10.1016/j.aquatox.2005.09.009 Grosse Y, Baan R, Straif K, Secretan B, El Ghissassi F, Bouvard V, Benbrahim L, Guha N, Galichet L, Cogliano V (2009) A review of human carcinogens—Part A: pharmaceuticals. Oncology 10:13-14. doi: 10.1016/S1470-2045(08)70286-9 Guillen D, Ginebreda A, Farré M, Darbra RM, Petrovic M, Gros M, Barceló D (2012) Prioritization of chemicals in the aquatic environment based on risk assessment: Analytical, modeling and regulatory perspective. Sci Tot Environ 440:236-252. doi:10.1016/j.scitotenv.2012.06.064 Helwig K, Colin H, McNaughtan M, Roberts J, Pahl O (2016) Ranking prescribed pharmaceuticals in terms of environmental risk: Inclusion of hospital data and the importance of regular review. Environmental Toxicology and Chemistry 35(4):1043–1050 Huijbregts MAJ, Hauschild M, Jolliet O, Margni M, McKone T, Rosenbaum RK, van de Meent D (2010a) USEtoxTM User manual. Available online at: http://www.usetox.org/sites/default/files/supporttutorials/user_manual_usetox.pdf. Accessed on September 28, .2014 Huijbregts MAJ, Margni M, van de Meent D, Jolliet O, Rosenbaum RK, McKone T, Hauschild M (2010b) Chemical-specific database: organics. Avalilable online at: http://www.usetox.org/sites/default/files/support-tutorials/database_organics.pdf..Accessed on September 28, 2014 Huijbregts MAJ, Struijs J, Goedkoop M, Heijungs R, Hendriks AJ, van de Meent D (2005a) Human population intake fractions and environmental fate factors of toxic pollutants in life cycle impact assessment. Chemosphere 61:1495–1504. doi: 10.1016/j.chemosphere.2005.04.046 Huijbregts MAJ, Rombouts LJA, Ragas AMJ, van de Meent D (2005b) Human-Toxicological Effect and Damage Factors of Carcinogenic and Noncarcinogenic Chemicals for Life Cycle Impact Assessment. Integr Environ Assess Manag 1(3):181-244 Iannacone J, Alvariño L (2009) Aquatic risk assessment of seven pharmaceutical products on Daphia magna. Ecol Apl 8(2):71-80. ISSN 1726-2216. ISO, 2006. ISO 14044:2006: Environmental Management - Life Cycle Assessment - Requirements and Guidelines. Geneva, Switzerland. Kobayashi Y, Peters GM, Khan SJ (2015) Towards more holistic environmental impact assessment: Hybridisation of Life Cycle Assessment and Quantitative Risk Assessment. Procedia CIRP 29:378-383 doi: 10.1016/j.procir.2015.01.064. Kumar A, Xagoraraki I (2010) Pharmaceuticals, personal care products and endocrine-disrupting chemicals in U.S. surface and finished drinking waters: A proposed ranking system. Sci Tot Environ 408:5972- 5989. doi:10.1016/j.scitotenv.2010.08.048 Larsson DGJ, de Pedro C, Paxeus N (2007) Effluent from drug manufactures contains extremely high levels of pharmaceuticals. Journal of Hazardous Materials 148:751–755. doi:10.1016/j.jhazmat.2007.07.008 Larsson DGJ (2014a) Antibiotics in the environment. UPS J Med Sci 119(2):108-112. doi: 10.3109/03009734.2014.896438 Larsson DGJ (2014b) Pollution from drug manufacturing: review and perspectives. Phil. Trans. R. Soc. B 369: 20130571. http://dx.doi.org/10.1098/rstb.2013.0571 Li D, Huijbregts MA, Jolliet O (2015) Life cycle health impacts of polycyclic aromatic hydrocarbon for source-specific mixtures. Environmental health Science 20(1):87-99 Lorenzo-Toja Y, Alfonsín C, Amores MJ, Aldea X, Marin D, Moreira MT, Feijoo G (2016) Beyond the conventional life cycle inventory in wastewater treatment plants. Sci Tot Environ 553:71-82. doi: 10.1016/j.scitotenv.2016.02.073