key: cord-0743135-5veiqexk
authors: Kosma, Christina I.; Kapsi, Margarita G.; Konstas, Panagiotis-Spyridon G.; Trantopoulos, Epameinondas P.; Boti, Vasiliki I.; Konstantinou, Ioannis K.; Albanis, Triantafyllos A.
title: Assessment of multiclass pharmaceutical active compounds (PhACs) in hospital WWTP influent and effluent samples by UHPLC-Orbitrap MS: Temporal variation, removals and environmental risk assessment
date: 2020-08-30
journal: Environ Res
DOI: 10.1016/j.envres.2020.110152
sha: e4fb1703f37a816e1d0911caaf431dc6a7e6c2a9
doc_id: 743135
cord_uid: 5veiqexk

Nowadays the occurrence and associated risks of Pharmaceutical Active Compounds (PhACs) in the aquatic environment comprises a major issue. In the present study, a comprehensive survey on contamination profiles, occurrence, removals, temporal variation and ecological risk of multiclass multiresidue PhACs, such as antibiotics, non-steroidal anti-inflammatories, lipid regulators and phsychiatrics, (including past and newly monitored PhACs as well as some of their metabolites) was performed in wastewaters from the WWTP of Ioannina University hospital along one year period on a monthly sampling basis. WWTP influent and effluent samples were analyzed for physicochemical quality parameters and PhACs concentration levels using Ultra High Performance Liquid Chromatography-Orbitrap-Mass Spectrometry (UHPLC-Orbitrap-MS), after Solid Phase Extraction (SPE) through Oasis HLB cartridges. Influent concentrations ranged between <LOQ (Limit of Quantification) for diclofenac and tolfenamic acid and 48586 ng/L for caffeine, while effluent concentrations between <LOQ for tolfenamic acid and simvastatin and 3361 ng/L for caffeine. Removal efficiencies ranged between -132.6% for venlafaxine and 100% for caffeine. Environmental risk assessment by means of Risk Quotient (RQ) for maximum and minimum concentration levels as well as optimized by the frequency of exceeding toxicity threshold values, RQ(f), was applied revealing that up to 12 PhACs posed acute toxicity (clofibric acid, fenofibrate, sulfadiazine, sulfamethoxazole, trimethoprim, amitryptiline, fluoxetine, fluvoxamine, norfluoxetine, sertraline, venlafaxine, caffeine) while up to 4 compounds exerted long-term toxicity (sulfamethoxazole, fluoxetine, sertraline, caffeine) at least for one of the studied organisms. Furthermore, mixture RQ(MEC/PNEC) and RQ(STU) effect of multiple compounds showed high potential risks of the target groups in some cases, although some contaminants were not included due to lack of available data. Results can be used to prioritization of PhACs and their metabolites for surveillance in receiving water bodies as well as development of knowledge on toxicity and mechanism(s) of action.

The last years numerous studies have been published concerning the presence of pharmaceutical active compounds (PhACs) in the aquatic environment. The extensive and long-term use of PhACs in human and veterinary medicine, due to the population growth and the development of medical techniques, has rendered them as contaminants of emerging concern (ECs) since they pose a potentially threat to the ecology of the receiving environment (Letsinger et al., 2019; Pereira et al., 2015; Wang et al., 2018) . The occurrence of PhACs in the environment becomes more concerning considering the fact that they do not exist alone but in mixtures which might lead into undesirable additive or synergistic effects (Kosma et al., 2014; Petrie et al., 2016) . The Institute for Health Informatics (IMS), reported that global medicine spending exhibits an increase of 4-7% in the past five years, depending on the country (Aitken, 2021) . In addition, more than 50% of all medicines are not prescribed, dispensed or sold in a proper way while many people cannot take them correctly (Fekadu et al., 2019) .

The main source of pharmaceutical pollution is highly associated with urban and hospital Wastewater Treatment Plant (WWTP) effluents (Krakkó et al., 2019) . Due to J o u r n a l P r e -p r o o f is the most useful tool in order to gain information about the variation of PhACs concentration, mainly in cases where there are not enough data regarding the consumption of them, as in the case of Greece (Papageorgiou et al., 2016) . Therefore, monitoring data concerning the occurrence and fate of PhACs during wastewater treatment processes is of crucial importance in order to evaluate the efficiency of the plants and PhACs contribution in the final receiving water body (Kosma et al., 2019) .

In the present work, a comprehensive monitoring study was performed in order to determine and evaluate the levels of a wide multiclass of PhACs in the University hospital WWTP of Ioannina city, located in northwestern Greece. In particular, a one-

year study took place in order to: (1) characterize the occurrence (levels, temporal variation and correlations) of 35 multiresidue multiclass PhACs in influent and effluent samples, by means of UHPLC-LTQ/Orbitrap HRMS. The selection of PhACs was based on their consumption, previous studies conducted by the same research group (Kosma et al., 2010; 2014; and was enriched by newly emerging PhACs as well as metabolites;

(2) to estimate the removal efficiencies and seasonal variation of past and newly emerging compounds and metabolites in the WWTP; (3) to provide an updated and completed environmental risk assessment of the target compounds in the aquatic environment. The hospital WWTP effluents were channeled into the municipal WWTP, which discharges its treated effluents into Kalamas River (the largest river of the region), being a considerable inflow to the River's flow, and eventually to its estuaries and the Ionian Sea (Kosma et al., 2019). Physicochemical parameters of influent and effluent wastewater samples during the sampling campaign are determined by standard methods, as described in supporting information section, and summarized in Table 1 (min, max, standard deviation (SD) and median values), Table S2 and Table S3 Extraction of PhACs in the influents and the effluents was carried out with Solid Phase Extraction (SPE), by using Oasis HLB (200 mg, 6 cm 3 ) cartridges according to the methodology described in supporting information section.

A UHPLC Accela LC system, connected with a hybrid LTQ-FT Orbitrap XL 2.5.5 SP1 mass spectrometer equipped with an electrospray ionization source (ESI) (Thermo Fisher Scientific, Inc., GmbH, Bremen, Germany) was used for the analysis of the samples. Full scan in positive (PI) and negative (NI) ionization mode was applied with mass resolving power of 60,000 FWHM and mass range of 100-500 Da and extracted ion chromatograms were used for identification and quantification purposes. In addition, a data-dependent acquisition (full MS/dd-MS 2 ) based on Collision Induced Dissociation (CID), was performed (Table S4) . Detailed operational parameters of LTQ-FT Orbitrap instrument was reported in supporting information section. Chromatographic separation was performed on a reversed phase SpeedCore C18 analytical column with 50 mm x 2.1 mm, 2.6 µm particle size (Fortis Technologies, Cheshire, UK). In PI 0.1% formic acid in water and 0.1% formic acid in methanol, used as mobile phases A and B, respectively. The elution gradient started at 95% A (initial conditions), remained 95% for 1 min, progressed to 30% in 3 min, then progressed to 0 % in 6 min, and returned to the initial conditions after 3 min with re-equilibration of the column set at 1 min. Total run time was 10 min. The flow rate was 0.4 mL/min with an injection volume of 5 µL. In NI water and methanol, used as J o u r n a l P r e -p r o o f mobile phases A and B, respectively. The elution gradient started at 90% A (initial conditions), remained 90% for 0.5 min, progressed to 30% in 2 min, reaching 10% in 3 min, decreased in 5% at 3.9 min, decreased in 0% at 4.5 min, remained 0% for 0.5 min, and returned to the initial conditions after 1 min with re-equilibration of the column set at 2 min. Total run time was 8 min. The flow rate was 0.4 mL/min with an injection volume of 20 µL. In both cases the analytes were separated at 27 ˚C.

Quality assurance and quality control (QA/QC), as well as validation studies in wastewater samples were based on the guidelines set by the EU Commission (European Commission Decision, 2002) . Quantification of PhACs was carried out by using matrix matched standard calibration curves (influent and effluent) prepared in 20:80 (v/v) methanol: water with 0.1% formic acid. In the case of psychiatric drugs matrix matched standard calibration curves contained the appropriate amount of each isotopically labeled standard (ILIS), as it was the only group of analytes for which ILIS were available (see Table S5 ). Since it was not possible to use the corresponding of the target compounds in wastewater ranged in the influents between 0.9 ng/L for CBZ and 45.6 ng/L for DMS and in the effluents between 0.5 ng/L for CBZ and 32.3 ng/L for DMS, respectively (Table S5 ). The linearity of the proposed method was investigated by constructing a 12-point calibration curve from LOQ to approximately 100 LOQ for each compound, applying correlation coefficient of ≥0.99, for all PhACs. The precision of the method was determined by the repeatability and reproducibility studies, and expressed as the relative standard deviation RSD (%). The RSD (n=5) values for intra-day ranged between 2.8 and 12.0% and the RSD for interday (n=5) ranged between 3.9 and 15.6% (Table S5) . Recoveries (n=3) were carried out in three spiking levels; 50, 250 and 500 ng/L (Table S6) . Matrix effect (signal suppression or enhancement) was calculated by applying the equation reported in our previous works (Kosma et al., 2014; 2015; and is given in Table S5 .

Removal efficiencies of PhACs were calculated as percentages of reduction between the concentration of PhACs in influent and effluent (Kosma et al., 2014; 2015; PNEC has been determined: (i) for acute toxicity, using the lowest values of EC50 or LC50 divided by an assessment factor (AF) of 1000 and (ii) for chronic toxicity, using NOEC divided by an AF of 100, 50 and 10 when one long-term NOEC, two-long term NOECs or three long-term NOECs, from the three aquatic organisms are available, respectively (European Commission, 2003; Kosma et al., 2014; . In the present study, an extensive literature review concerning ecotoxicological information for fish, invertebrates and algae was conducted, for acute and chronic toxicity, and the data are given in Tables S7 and S8. Levels of concern taken into consideration for RQ were as follows; high ecological risk for RQ≥1, medium risk for 0.1<RQ<1 and low or negligible risk for 0.01<RQ<0.1 (Hernando et al., 2006; Kosma et al., 2014) .

Furthermore, according to Zhou et al. (2019) the optimized risk quotient value (RQ f ) was determined based on mean RQ value and the frequency of MECs exceeding PNEC. Estimation of RQ f by considering the variability of concentration above PNEC favors to screen the pollutants that are widely distributed and frequently detected. The RQ f was calculated according to the following equation (Eq. (1)): Input datasets were standardized, namely they have been transformed by subtracting some reference value (typically a sample mean) from each value and dividing it by the standard deviation (typically a sample SD). Standardization makes the results of a variety of statistical techniques entirely independent of the ranges of values or the units of measurements. Pearson's moment correlation factor (r) was performed between the variables, and those strongly correlated were unselected from further statistical analysis to avoid multicollinearity. For data analysis, concentration values were set to half of LOD for not detected (below LOD) cases and half of the LOQ if they were between LOD and LOQ.

J o u r n a l P r e -p r o o f Fig. S1 and Table S9 . Further discussion for each therapeutic category is given in the following subsections.

Among the analgesics/NSAIDs, the metabolite SA was detected in 100% and 66.7% of influent and effluent samples, respectively. In influent samples concentration levels ranged from 24.9 ng/L in August to 3603 ng/L in April while in the effluents from 15.2 ng/L in September to 64.8 ng/L in January. SA is the major metabolite of acetylsalicylic acid (aspirin ® ) but there are several other sources from which it can derive, since it is also used as an additive in some skin-care products, in toothpastes and as a food preservative (Konstas et al., 2019) . In Greece, aspirin is one of the most popular first line analgesics while it is also a freely available over-the-counter drug (Papageorgiou et al., 2016) . SA was detected during the whole year in the influents with higher mean concentrations in winter (2245.3 ng/L) and lower in summer (122.9 ng/L), although with no significant differences (p>0.05; p=0.108), since it constitutes an NSAID which is consumed also to cure inflammation caused by influenza, which occurs more frequently during colder months (Tang et al., 2019) . As for the other two analgesics/anti-inflamatories, DCF and TA concern, their detected concentrations were only at levels <LOQ. ΤΑ is excreted unchanged in 8.8%, therefore, low concentrations might enter the influent, while for DCF excretion rates vary from <1- 

Regarding lipid regulators, FNB was the most frequently detected compound, 

Antibiotics are one of the most frequently detected therapeutic groups in hospital effluents, mainly due to their excretion in urine (up to 63%) and, in a less extent, in faeces, as metabolites or in the unchanged form. Sulfonamides (SAs) are among the most ubiquitous classes of antibiotics found in hospital effluents due to their characteristics (broad-spectrum, low price and stable properties) together with other antibiotics like TMP. Usually, they are found at higher concentrations than those reported in urban wastewaters, reaching a few hundreds of µg/L (Verlicchi, 2018).

According to Zhang et al. (2020) the overuse and abuse of antibiotics have promoted the development and spread of antibiotic resistance genes, originating from antibiotic resistant bacteria, which pose potential risks to human health and eco-environment. In the present study, detection of four SAs and TMP was investigated. In the influents, SDZ, SMX, SPY and STZ were detected in 66.7, 91.7, 41.7 and 8.3% of the samples. Furthermore, it was noticed that effluent concentrations of antibiotics in some cases J o u r n a l P r e -p r o o f were higher than the influents, which could be ascribed to the conversion of the conjugated compounds to their parent during the treatment process (Cui et al., 2020) .

Prescription of psychiatric drugs during the last years have been increased, therefore Many psychiatric compounds have been detected at similar or higher concentrations in the effluents than in the influents. This is might be due to the microbial transformation of parent or conjugated forms of these drugs through enzymatic processes that take place in the WWTP or changes in the adsorption behavior to particles during the treatment processes (Kosma et al., 2019). PCs with eigenvalues >1 is sufficiently explaining most of the variance and those PCs were kept to analysis (Table S12 ).

J o u r n a l P r e -p r o o f the selected emerging contaminants reveals certain differences. In influent, PC1, which is a linear combination of the sum of means of the psychiatrics, stimulants and antibiotics, accounted for the 38.92% of the total variability (Table S12) In both plots (Fig. 3) , red dots are the twelve months (cases) of sampling during the one-year campaign. Concerning influent, similarities between cases according to their point markers distances reveal common characteristics for months starting at the end of spring till the beginning of autumn including May, June, July, August and

September. This is consistent taking into consideration the concentrations detected in influent during that time period and the common weather characteristics exhibited in Greece. On effluent PCA plot, one could assume that September is quietly in distance J o u r n a l P r e -p r o o f with almost all other months. It has to be noted that in both plots, September's contribution to PC1 variance is executive and that is due to the higher concentrations detected in influent and effluent that month of year.

In Fig. S2 (A,B) 

Concentration heat map (Fig. 4) shows temporal (monthly) variations of target PhACs. In the influent, the higher concentrations variation (red color) of individual drugs is mainly located in months September, October and July. Taking into consideration the fact that in July, the presence of two main psychiatric drugs' metabolites is ubiquitous, one could assume also that the exceptional weather Clustering of the 35 PhACs (vertical axis, Fig. 4 ), on the other hand, is more complicated in both influent and effluent. In effluent, a distinct cluster at first, separates almost half of the detected analytes but the next grouping consists mainly of psychiatric drugs, two of their metabolites and FNB along with BZF. Other distinct clusters exist as for example, SMX and TMP grouping, which are known to be used together. In effluent, the first distinct cluster contains psychiatric drugs, as RIS, AMT, FLX, AMS and CIT as well as antibiotics such as SMX, SDZ and TMP; FNB and CA (lipid regulators) and DCF (analgesics).

Removal efficiencies were evaluated according to our previous works ( reported that one major mechanism which is associated with the biodegradation of PhACs in the WWTP is the co-metabolism, where the compound is degraded through the enzymes secreted by the microorganisms of the biological sludge. It was found that for some PhACs, including BZF, the co-metabolic biodegradation process was the main mechanism involved in its removal process (Couto et al., 2019) .

The removals of antibiotics varied also from negative in many cases (-33% for SMX)

to high values (99.5% for SPY). SAs and particularly SMX presented in most of the cases negative removals. It was found that aromatic compounds bearing sulphate or halogen functional group are not easily degraded by biological treatments (Couto et al., 2019) . Furthermore, it has been found that degradation of ionizable compounds such as SMX is highly depended on the pH. In acidic media SMX is in a hydrophobic Removals' seasonal variation was also estimated and is depicted in Fig. S4 Concerning acute toxicity, high risks (RQ max >1) were posed in invertebrates by CA, TMP, FLX, STR, VNX and CAF. CA presented also high RQ mean and RQ f in invertebrates (RQ fCAF =2.303, F=16.7%). This fact, indicates that there is a potential high risk of CA, but the possibility for the organisms to be exposed to unsafe levels is Verlicchi (2018) reported that antibiotics were identified as the therapeutic group present in hospital effluents that cause the greatest concern and they were pointed out as the main contributors for the high environmental risk of these effluents. In addition, algae showed to be the most sensitive species to the toxic effect of antibiotics, nevertheless antibiotics might pose risk to all the trophic levels, representing a threat for the entire aquatic ecosystem. In the same study, it was referred that RQs calculated for hospital wastewaters varied widely from 10 -3 to 10 3 . Frédéric and Yves (2014) reported RQ >1 for TMP, and SMX, while Mendoza et al. (2015) found RQ>1 for TMP in a hospital from Valencia (Spain). Furthermore, it was found that measured RQ was higher than 1.0 for SMX taking into consideration urban and hospital consumptions. As a result, SMX was identified as high priority compound for the water cycle (Verlicchi, 2018) . Felis et al. (2020) reported that harmful effects of antibiotics in algae might cause disorders to the organisms of higher trophic levels.

Since algae are at the bottom of the food chain, even small changes in the algal population may dis-equilibrate the aquatic system (Lin et al., 2018) . Concerning RQs for the mixture of PhACs, Table 3 represents the RQ MEC/PNEC , RQ STU and their ratio, for acute and chronic toxicity, taking into consideration maximum and mean concentrations, respectively. It is worth noticing that in some cases available data concerning acute and chronic toxicity were missing, thus the corresponding RQs 

The occurrence, removal, temporal variation and ecological risk of 35 PhACs, belonging to various therapeutic categories, in a University hospital WWTP were investigated monthly, along one year. Concentrations found at levels from a few ng/L to a few µg/L. In the influents all monitored compounds were detected at Environmental impact of pharmaceuticals from Portuguese wastewaters:

Geographical and seasonal occurrence, removal and risk assessment. Table 3 Calculated RQ MEC/PNEC , RQ STU and their ratio, based on acute and chronic toxicity for mean and maximum concentrations. Factor 1 : 38.92% 

Outlook for Global Medicines through

Predictive Environmental Risk Assessment of Chemical Mixtures

Screening level mixture risk assessment of pharmaceuticals in STP effluents

Caffeine , an Anthropogenic Marker for Wastewater Contamination of Surface Waters

Occurrence, fate and removal of pharmaceutically active compounds (PhACs) in water and wastewater treatment plants-A review

Occurrence and fate of psychiatric pharmaceuticals in the urban water system of Shanghai

Detection, occurrence and fate of 22 psychiatric pharmaceuticals in psychiatric hospital and municipal wastewater treatment plants in Beijing

Efficient multiresidue determination method for 168 pharmaceuticals and metabolites: Optimization and application to raw wastewater, wastewater effluent, and surface water in Beijing

Optimization of screening-level risk assessment and priority selection of emerging pollutants -The case of pharmaceuticals in European surface waters

This research was co-funded by the European Union and National Funds of the participating countries (Interreg-IPA CBC, Greece-Albania, "PhaRem"). The authors would like to thank the Unit of Environmental, Organic and Biochemical highresolution analysis-Orbitrap-LC-MS of the University of Ioannina for providing access to the facilities.

• Removal efficiencies ranged between -132.6% and 100%.• Risk assessment was applied for single compounds and observed mixture cases.• Optimized RQs calculated using PhAC mean concentration and detection frequency.

x The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.☐The authors declare the following financial interests/personal relationships which may be considered as potential competing interests: J o u r n a l P r e -p r o o f