Occurrence of Iodinated X-ray Contrast Media and Their

Aug 31, 2011 - Biotransformation Products in the Urban Water Cycle. Jennifer Lynne Kormos, Manoj Schulz, and Thomas A. Ternes*. Federal Institute of ...
3 downloads 0 Views 3MB Size
ARTICLE pubs.acs.org/est

Occurrence of Iodinated X-ray Contrast Media and Their Biotransformation Products in the Urban Water Cycle Jennifer Lynne Kormos, Manoj Schulz, and Thomas A. Ternes* Federal Institute of Hydrology (BfG), Am Mainzer Tor 1, D-56068, Koblenz, Germany

bS Supporting Information ABSTRACT: A LC tandem MS method was developed for the simultaneous determination of four iodinated X-ray contrast media (ICM) and 46 ICM biotransformation products (TPs) in raw and treated wastewater, surface water, groundwater, and drinking water. Recoveries ranged from 70% to 130%, and limits of quantification (LOQ) varied between 1 ng/L and 3 ng/L for surface water, groundwater and drinking water, and between 10 ng/L and 30 ng/L for wastewater. In a conventional wastewater treatment plant, iohexol, iomeprol, and iopromide were transformed to >80%, while iopamidol was transformed to 35%. In total, 26 TPs were detected above their LOQ in WWTP effluents. A significant change in the pattern of ICM TPs was observed after bank filtration and groundwater infiltration under aerobic conditions. Predominately, these TPs are formed at the end of the microbial transformation pathways in batch experiments with soil and sediment. These polar ICM TPs, such as iohexol TP599, iomeprol TP643, iopromide TP701A, and iopromide TP643, were not or only partially removed during drinking water treatment. As a consequence, several ICM TPs were detected in drinking water, at concentration levels exceeding 100 ng/L, with a maximum of 500 ng/L for iomeprol TP687.

’ INTRODUCTION Iodinated X-ray contrast media (ICM) are used for the imaging of internal body structures (i.e., organs, blood vessels, and soft tissues) during diagnostic examinations.1 ICM are reported to be the most widely used pharmaceuticals for intravascular administration and most frequently used in hospitals.1,2 They are applied at high doses (i.e., up to 200 g/application) and are eliminated nonmetabolized in the urine within 24 h.3 Most ICM are derivatives of 2,4,6-triiodobenzoic acid and are classified as ionic or nonionic depending on the functional moieties at their side chains. For instance, the ionic ICM, diatrizoate, is negatively charged at neutral pH due to its carboxylic moiety (pKa = 3.4), while the nonionic ICM (i.e., iohexol, iopamidol, and iomeprol) contain functional moieties which are uncharged at neutral pH.4,5 It has been reported in several studies that these hydrophilic and metabolically stable ICM are not effectively removed in conventional wastewater treatment plants (WWTPs).68 The elevated concentrations of ICM in surface water, groundwater, and even treated water of drinking water treatment plants (DWTPs) can be explained by their persistence.2,913 However, recent research has shown that certain oxidation processes (i.e., ozonation, advanced oxidation processes) as well as biological processes of WWTPs with elevated sludge retention times (SRT) of >12 d are capable of transforming nonionic ICM.1417 Nevertheless, elimination did not result in mineralization of the parent ICM. In most cases, oxidation products or biotransformation products (TPs) were formed.16,1822 Aerobic r 2011 American Chemical Society

biotransformation reduced the length of the ICM side chains, while electrochemical and photochemical processes may lead to deiodinated products.17,23 Recently, 46 TPs of four nonionic ICM have been identified which were grouped into different phases depending on the sequence of their formation.2022 So far, minimal efforts have focused on the presence of these newly identified TPs in the aquatic environment and drinking water supplies. The main objective of this study was to investigate the occurrence and fate of the newly identified TPs of four nonionic ICM (iohexol, iomeprol, iopamidol, and iopromide) in different aqueous matrixes using an optimized LC tandem MS analytical method. In particular, this study focuses on changes occurring in the composition and patterns of these TPs in the urban water cycle from WWTPs via surface water to groundwater and drinking water.

’ EXPERIMENTAL SECTION Description of Sampling Locations. Wastewater Treatment Plant (WWTP). Five-day composite samples were collected from a

conventional German WWTP during dry weather conditions in December 2009. This WWTP serves 285 000 population equivalents (PE), has a SRT and hydraulic retention time (HRT) of Received: December 18, 2010 Accepted: August 11, 2011 Revised: August 1, 2011 Published: August 31, 2011 8723

dx.doi.org/10.1021/es2018187 | Environ. Sci. Technol. 2011, 45, 8723–8732

Environmental Science & Technology

ARTICLE

Figure 1. Flow scheme of a conventional German WWTP consisting of mechanical and biological treatment. Sampling points 1 and 2 refer to the influent 1 and 2; sampling point 3 refers to prebiological treatment; sampling point 4 refers to postbiological treatment; and sampling point 5 refers to final effluent.

16 d and 60 h, respectively, and an average wastewater inflow of approximately 35 500 m3/d. It receives two different wastewater streams from the surrounding area, consisting of domestic sewage from various districts of the city as well as hospital wastewaters (i.e., University hospital). This WWTP applies mechanical treatment, biological phosphate removal, denitrification, nitrification, secondary sedimentation, and filtration. The mechanical treatment consists of different sized screens, a grit removal tank (sand filtration tank), and a primary sedimentation tank. After primary sedimentation, the wastewater enters a biological phosphate elimination tank, followed by an activated sludge tank containing denitrification and nitrification. The effluent enters a secondary sedimentation tank and a two-layer filter (consisting of anthracite and sand) before being discharged into the receiving water body. Figure 1 provides a schematic diagram of this WWTP and the sampling sites. Samples were taken from both influent streams (sampling point 1 and 2), prior to denitrification and nitrification (sampling point 3), after secondary sedimentation (sampling point 4), and after filtration (sampling point 5). Rhine River Water and Riverbank Wells. On November 18th, 2009, three riverbank wells close to the Rhine River were sampled at Urmitz, a small town in Germany close to Koblenz. The distance from the wells to the river ranged from 36 to 150 m, and the water level depths varied between 8.53 and 9.29 m. In addition, four-week composite samples were collected along the Rhine River at Koblenz, a distance of 590.3 km upstream from Lake Constance, from September to November 2009. Water quality parameters of the individual wells are summarized in Table S1 of the Supporting Information. The river bank and soil surrounding the wells consisted of crushed rock and gravel. The average infiltration velocity from the Rhine River into groundwater was approximately 13 m/d. DWTP Treatment Processes. Water was collected from four German DWTPs (DWTP14) after individual treatment processes. The raw water source for DWTP1 is a mixture of groundwater and bank filtrate, which is then directed to a granular activated carbon (GAC) filtration system before being distributed. DWTP2 receives its raw water from bank filtration of surface water, artificial groundwater recharge, and natural groundwater. Aeration and sand filtration are the main treatment processes which are capable of removing contaminants. DWTP3 uses river water as a raw water source and is treated by flocculation with Fe(III)chloride, followed by sedimentation, ozonation, multilayer filtration, and then GAC filtration. Samples were collected from the raw water source as well as after

ozonation and after GAC filtration. DWTP4, applying a similar multibarrier approach as DWTP3, uses a combination of river water and groundwater as raw water sources. After the river water undergoes flocculation with Fe(III)chloride, sedimentation, sand filtration, and GAC filtration, it is mixed with groundwater and is treated by aeration slow sand filtration and chlorine dioxide disinfection. Samples were collected from the river, after GAC filtration of the river water, after mixing of the treated water and groundwater, and finally after the addition of chlorine dioxide. Sample Preparation and Extraction for ICM TPs. The aqueous samples were filtered through glass fiber filters (Schleicher and Schuell, Dassel, Germany), acidified to pH 2.6 by adding sulfuric acid (3.5 M H2SO4), and stored at 4 °C prior to extraction. Two solid phase extraction (SPE) cartridges, C18 cartridge (3 mL, 200 mg, J.T. Baker) placed on top of Bakerbond SDB-1 (3 mL, 200 mg, J.T. Baker), were used for cleanup and extraction, respectively. Both SPE cartridges were conditioned with 4  1 mL of methanol and 4  1 mL of groundwater adjusted to pH 2.6 with 3.5 M H2SO4. For raw and treated wastewater, 100 and 200 mL were extracted, respectively, and 500 mL or 1 L was extracted for surface water, groundwater, and drinking water. Prior to extraction, all samples were spiked with 10 μL of a surrogate solution (20 ng/μL). The surrogate standard solution consisted of iohexol-d5, iomeprol-d3, iopamidol-d3, diatrizoate-d6, and desmethoxyiopromide (DMI). DMI was kindly provided by Bayer Schering Pharma (Berlin, Germany), diatrizoate-d6 was purchased from Campro Scientific (Berlin, Germany), and iohexol-d5, iomeprol-d3, and iopamidol-d3 were purchased from Toronto Research Chemicals (North York, Canada). After enrichment by SPE, the cartridges were washed with 4 mL of Milli-Q water adjusted to pH 2.6 by adding 3.5 M H2SO4. The C18 cartridges were disposed of, and the SDB-1 cartridges were dried under a gentle stream of nitrogen gas. The analytes bound on the sorbents in the SDB-1 cartridges were eluted with 4  2 mL of methanol, evaporated to 100 μL using nitrogen gas, and reconstituted with 900 μL of Milli-Q water. The extracted samples were stored at 4 °C until measured by LC tandem MS. LC ESI(+) Tandem MS Detection of ICM and TPs. An Agilent 1200 Series HPLC system consisting of an autosampler, binary pump, degasser (Waldbronn, Germany), and a column oven (MayLab Analytical Instruments, Austria) was used. Chromatographic separation of the analytes was achieved on two coupled Chromolith Performance RP-18e columns (4.6 mm  100 mm, Merck, Darmstadt, Germany) equipped with a Chromolith 8724

dx.doi.org/10.1021/es2018187 |Environ. Sci. Technol. 2011, 45, 8723–8732

Environmental Science & Technology

ARTICLE

Table 1. Recoveries (%) and 95% Confidence Intervals (n = 3) for ICM and Selected TPs in the Investigated Environmental Matrixes Analyte Iohexol Iohexol TP687A Iohexol TP657 Iohexol TP599 Iomeprol Iomeprol TP701

WWTP influenta

WWTP effluenta

River waterb

Groundwaterb

Drinking waterc

91 ( 25

120 ( 21

71, 89 (n = 2)

106 ( 14

96 ( 31

95 ( 11 106 ( 21

119 ( 6 117 ( 15

78 ( 12 90 ( 10

75 ( 12 105 ( 4

84 ( 10 120 ( 7 136 ( 18

83 ( 49

123 ( 36

82 ( 5

85 ( 2

123 ( 158d

92 ( 21

80 ( 9

113 ( 14

89 ( 41

94 ( 29

108 ( 22

121 ( 26

106 ( 29

96 ( 27 120 ( 40

Iomeprol TP643

90 ( 27

104 ( 15

84 ( 27

104 ( 46

Iomeprol TP629

105 ( 17

112 ( 9

74 ( 8

82 ( 16

83 ( 26

Iopamidol

111 ( 13

100 ( 9

117 ( 18

113 ( 20

128 ( 38

Iopamidol TP791 Iopamidol TP761

87 ( 4 112 ( 19

93 ( 5 93 ( 21

90 ( 17 88 ( 14

93 ( 6 97 ( 10

87; 86 (n = 2) 141 ( 29

Iopamidol TP745

94 ( 4

99 ( 3

88 ( 14

78 ( 6

109 ( 25

Iopromide

79 ( 13

89 ( 14

119 ( 33

105 ( 1

111 ( 15

Iopromide TP819

78 ( 4

91 ( 2

87 ( 21

41 ( 14

111 ( 15

Iopromide TP31B

86 ( 12

97 ( 3

115 ( 24

84 ( 11

101 ( 12

Iopromide TP729A

121 ( 14

112 ( 15

137 ( 13

110 ( 19

120 ( 16

Iopromide TP759

82 ( 10

95 ( 10

102 ( 3

93 ( 15

123 ( 18

Iopromide 701A Iopromide 701B

81 ( 4 78 ( 9

92 ( 3 86 ( 4

100 ( 15 114 ( 16

94 ( 13 101 ( 16

111 ( 16 105 ( 11

Iopromide TP643

83 ( 11

99 ( 10

76 ( 13

72 ( 18

71 ( 13

Spiking concentration of 50 μg/L for ICM concentrations higher than 10 μg/L (in those cases direct injection was used without SPE). b Spiking concentration of 0.1 μg/L. c Spiking concentration of 0.05 μg/L. d Spiking concentration was too low (background concentrations were high compared to spiking concentration). a

RP-18e guard column (4.6 mm  5 mm, Merck, Darmstadt, Germany). A sample aliquot of 50 μL was injected into the LC tandem MS, and the analytes were eluted from the column using two mobile phases, 95% Milli-Q water, 5% methanol, and 0.5% formic acid (A) and 99.5% methanol and 0.5% formic acid (B). The gradient elution program was as follows: 02 min, 100% A; 17 min, 90% A; 17.120 min, 100% A. Iopromide and its TPs were measured in a different HPLC run using an isocratic elution program which consisted of 90% A and 10% B for 20 min. A flow rate of 0.8 mL/min and column oven temperature of 50 °C was used for the measurements of ICM and TPs. The HPLC system was coupled to a 4000 Q Trap MS system (Applied Biosystems/MDS Sciex, Darmstadt, Germany) consisting of an electrospray ionization (ESI) source (operated in positive ionization mode). The source-dependent parameters were optimized for the ICM and their TPs. These parameters are summarized in Table S2 of the Supporting Information. Two mass transitions were optimized for the parent ICM and each TP for identification and confirmation purposes in MRM mode. The MRM transitions and compound-dependent parameters for the parent ICM and TPs as well as the surrogate standards are summarized in Table S3S7 of the Supporting Information. Method Validation. Recoveries for the analytical method were determined by spiking the analytes and surrogate standards into raw and treated wastewater, surface water, groundwater, and drinking water. The spiking concentrations were based on background concentrations of the parent ICM previously detected in different matrix types. For instance, raw and treated wastewater were spiked with concentrations of 0.1 and 1 μg/L, while surface water, groundwater, and drinking water were spiked at 0.05 and 0.1 μg/L. The recoveries were calculated according to eq 1, where Cspiked is the analyte concentration in the spiked sample,

Cnonspiked is the analyte concentration in the nonspiked sample, and Cinitial is the spiking concentration. Recovery ½% ¼

ðCspiked  Cnonspiked Þ  100 ðCinitial Þ

ð1Þ

Variations from the mean values (n = 3) are given as 95% confidence intervals. Calibration samples were prepared by spiking 10 μL of a surrogate standard solution (20 ng/μL), the standard solutions of iohexol, iomeprol, and iopamidol, and nine isolated TP standards (iohexol TP687A, TP657, TP599; iomeprol TP701, TP643, TP629; iopamidol TP791, TP761, TP745) as well as iopromide and its seven isolated TPs (TP819, TP759, TP731B, TP729A, TP701A, TP701B, TP643) into a final volume of 1 mL of Milli-Q water. A detailed explanation of the isolation of TPs and the preparation of the standard solutions can be found elsewhere.20,22 The 12 point calibration curve ranged from 1 to 3000 ng/L. Linear and quadratic regressions were applied to the calibration curves with a weighing factor of 1/x. The ICM TPs which could not be isolated in sufficient quantities so far were quantified using either a calibration curve of the parent compound or an isolated TP showing the same detected MS fragment ions. Since there was no guarantee that this procedure was always accurate, these TPs were marked with an asterisk indicating that the data is semiquantitative. A summary of the analytes as well as the surrogate standards used for quantification (semiquantification) of each analyte are summarized in Table S8S9 of the Supporting Information. In addition, samples obtained from the soilwater batch systems spiked with ICM (i.e., 1 g/L) according to Schulz et al.20 and Kormos et al.21 were analyzed to verify the correctness of the retention times of ICM TPs having similar MRM transitions. 8725

dx.doi.org/10.1021/es2018187 |Environ. Sci. Technol. 2011, 45, 8723–8732

Environmental Science & Technology

ARTICLE

Figure 2. Concentration (nmol/L) of iopromide and its TPs (a) and iomeprol and its TPs (b) detected prior to denitrification and nitrification (sampling point 3), after denitrification and nitrification (sampling point 4), and in the final effluent (sampling point 5) at the WWTP.

In brief, samples (10 μL) from the batch systems were taken at specific time intervals and spiked with 10 μL of the surrogate standard (20 ng/μL) solution and diluted to 1 mL with Milli-Q water prior to LC tandem MS detection. The limits of quantification (LOQ) were defined as a signal/ noise (S/N) ratio of >10, which was individually determined in each environmental sample and calibration sample for each analyte. In any case, the LOQ values were always higher than the lowest calibration point.

’ RESULTS AND DISCUSSION Method Validation for ICM and TPs in Aqueous Matrixes. A LC tandem MS method was developed to determine the

occurrence of four nonionic ICM (iohexol, iomeprol, iopamidol, and iopromide) and 46 ICM TPs of the nonionic ICM in various aqueous matrixes. The recoveries for the parent ICM and selected TPs are summarized in Table 1. In general, the recoveries varied between 70 and 130% with 95% confidence intervals (n = 3) less than 30%. The LOQ for analytes detected in surface water, groundwater, and drinking water ranged from 1 to 3 ng/L. The LOQ in raw and treated wastewater ranged from 10 to 30 ng/L due to elevated background matrix composition and lower sample volume of 100 and 200 mL, respectively. The analytical methods were able to quantify ICM and 46 ICM TPs in a variety of aqueous matrixes with sufficient accuracy and sensitivity. Transformation of ICM and TPs in Municipal WWTPs. In order to facilitate the description of the results for so many TPs, 8726

dx.doi.org/10.1021/es2018187 |Environ. Sci. Technol. 2011, 45, 8723–8732

7.1 ( 1.2

1400 ( 150

(mmol/d)

0.69 ( 0.24 (10 ( 3) 12 ( 4

2.1 ( 0.3

(28 ( 4)

36 ( 5

nmol/L

8727

4.3; 4.3

(130; 130) 150; 150

nmol/L

(g/d) (mmol/d)

336 ( 48

325 ( 65

150; 150

90 ( 9

(mmol/d)

Eliminationi %

97 ( 19

(68 ( 13)

2.9 ( 0.6

2.0 ( 0.4

(36 ( 7) 52 ( 10

1.5 ( 0.3

1.1 ( 0.2

16 ( 3

(12 ( 3)

0.42 ( 0.07

0.30 ( 0.05

4.4 ( 0.6

(3.2 ( 0.4)

0.26 ( 0.03

0.19 ( 0.02

5.8 ( 1.1

0.32 ( 0.06 (4.2 ( 0.8)

0.23 ( 0.04

Phase III TPsd

84 ( 5

580 ( 140

(450 ( 110)

17 ( 4

13 ( 3

(410 ( 110) 530 ( 17

16 ( 0.5

12 ( 0.4

3500 ( 80

(2700 ( 60)

91 ( 2

71 ( 2

123 ( 4

(95 ( 3)

7.3 ( 0.3

5.7 ( 0.2

3100 ( 300

168 ( 16 (2400 ( 230)

131 ( 12

Iomeprol

129 ( 34

462 ( 80

(300 ( 50)

14 ( 2 (110 ( 30)

8(2 3.8 ( 1.0

(170 ( 60) 256 ( 92

7.6 ( 2.7

5(2

61 ( 8

(40 ( 5)

1.6 ( 0.2

1.0 ( 0.1

13 ( 5

(9 ( 3)

0.79 ( 0.30

0.5 ( 0.2

19 ( 4

1.1 ( 0.2 (13 ( 3)

0.7 ( 0.1

Phase II TPsf

3.1 ( 0.9

(95 ( 37) 122 ( 46

3.6 ( 1.4

3(1

58 ( 19

(45 ( 15)

1.5 ( 0.5

1.2 ( 0.4

4.5 ( 0.4

(3.5 ( 0.3)

0.27 ( 0.02

0.21 ( 0.02

14 ( 4

0.74 ( 0.19 (11 ( 3)

0.6 ( 0.2

Phase I TPse

90 ( 4

48 ( 3

(40 ( 3)

1.4 ( 0.1

1.2 ( 0.1

(31 ( 7) 37 ( 8

1.1 ( 0.2

0.9 ( 0.2

500 ( 14

(410 ( 10)

13 ( 0.3

11 ( 0.3

76 ( 8

(62 ( 6)

4.5 ( 0.5

3.7 ( 0.4

404 ( 42

22 ( 2 (330 ( 30)

18 ( 2

Iohexol