Vous êtes sur la page 1sur 16

Wat. Res. Vol. 32, No. 11, pp.

3245±3260, 1998
# 1998 Elsevier Science Ltd. All rights reserved
Printed in Great Britain
PII: S0043-1354(98)00099-2 0043-1354/98 $19.00 + 0.00

OCCURRENCE OF DRUGS IN GERMAN SEWAGE


TREATMENT PLANTS AND RIVERS*
THOMAS A. TERNES{
ESWE-Insitute for Water Research and Water Technology, SoÈhnleinstrasse 158, D-65201 Wiesbaden,
Germany

(First received July 1997; accepted in revised form February 1998)

AbstractÐThe occurrence of 32 drug residues belonging to di€erent medicinal classes like antiphlogis-
tics, lipid regulators, psychiatric drugs, antiepileptic drugs, betablockers and b2-sympathomimetics as
well as ®ve metabolites has been investigated in German municipal sewage treatment plant (STP) dis-
charges, river and stream waters. Due to the incomplete removal of drug residues during passage
through a STP, above 80% of the selected drugs were detectable in at least one municipal STP e‚uent
with concentration levels up to 6.3 mg lÿ1 (carbamazepine) and thus resulting in the contamination of
the receiving waters. 20 di€erent drugs and 4 corresponding metabolites were measured in river and
stream waters. Mainly acidic drugs like the lipid regulators beza®brate, gem®brozil, the antiphlogistics
diclofenac, ibuprofen, indometacine, naproxen, phenazone and the metabolites clo®bric acid, feno®bric
acid and salicylic acid as well as neutral or weak basic drugs like the betablockers metoprolol, propra-
nolol and the antiepileptic drug carbamazepine were found to be ubiquitously present in the rivers
and streams, mostly in the ng lÿ1-range. However, maximum concentrations were determined up to
3.1 mg lÿ1 and median values as high as 0.35 mg lÿ1 (both beza®brate). The drugs detected in the en-
vironment were predominantly applied in human medicine. It can therefore be assumed that the load of
municipal STP e‚uents in the surface water highly in¯uences the contamination. Due to their wide-
spread presence in the aquatic environment many of these drugs have to be classi®ed as relevant en-
vironmental chemicals. # 1998 Elsevier Science Ltd. All rights reserved

Key wordsÐdrugs, antiphlogistics, lipid regulating agents, anticancer agents, diazepan, betablockers,
b2-sympathomimetics, carbamazepine, rivers and streams, sewage treatment plant e‚uents

INTRODUCTION behavior (half life, urinary and faecal excretion,


metabolism etc.).
Organic compounds, isolated from plants as well as
those synthesized by chemists, have been used as
Environmental exposure to drugs
drugs for a long time. To date many tons of drugs
are produced per year, and applied in human and Investigations in the United Kingdom as reported
veterinary medicine. Generally, exact production by Waggott (1981), Watts et al. (1983) and
®gures have not been published in the literature. Richardson and Bowron (1985) revealed that drugs
However, the amounts of pharmaceuticals pre- were present in the aquatic environment at concen-
scribed by physicians can be evaluated by multiply- trations up to approximately 1 mg lÿ1, whereas the
exact concentrations for the individual drugs were
ing the amount of daily dose with the number of
not always determined. Even in potable water and
prescribed daily doses per year. In Germany for
in ground water some drugs like diazepam, metha-
instance up to about 100 t of individual drugs
qualone and the antibiotics with penicilloyl groups
(Table 1) were prescribed in 1995 (Schwabe and have been determined. On Iona Island (Vancouver/
Pa€rath, 1995). This amount can increase dramati- Canada) the two antiphlogistics, ibuprofen and
cally when drugs are considered which can be pur- naproxen have been identi®ed in sewage (Rogers et
chased without prescription in the pharmacy. Due al., 1986). Hignite and Azarno€ (1977) detected
to this high application level, detectable concen- loads of salicylic acid up to 28.7 kg dÿ1 and of clo-
trations of drugs and their metabolites may be ®bric acid up to 2.9 kg dÿ1 in the e‚uents of the
expected in sewage. However, these concentrations municipal sewage treatment plant (STP) of Kansas
are dependent on the drugs pharmacokinetical City (USA). In Germany the clo®bric acid, a
metabolite of three lipid regulating agents, has been
identi®ed in river and ground water and even in
*Dedicated to Professor Dr. Klaus Haberer on the oc-
casion of his 70th birthday. drinking water with concentration levels ranging up
{[Tel: +611-780-4343; Fax: +611-780-4375, E-mail: ter- to 165 ng lÿ1 by Stan and LinkerhaÈger (1994),
nes@goofy.zdv.uni-mainz.de]. Heberer and Stan (1996). Our institute identi®ed a
3245
3246 Thomas A. Ternes

Table 1. Estimated German prescription amounts from human application in 1995 and estimated human excretion rates of some selected
drugs glucuronide-conjugates
Estimated prescription amounts in
Pharmacokinetical excretion rate in %* 1995 in tons per year$

Unchanged drug Glucuronides%

Beza®brate} 50 22 30
Clo®bric acid (clo®brate, 16
eto®brate, etophyllinclo®brate) 6 >90
Feno®bric acid (feno®brate) + +++ 15
Gem®brozil} ÿ 50 6
Diclofenac6 15 <1 75
Ibuprofen 1±8 14 105
Indometacine 10±20 80 6
Metoprolol 3±10 ÿ 50
Propranolol <1 ÿ 3
Carbamazepine** 1±2 + 80
+: small percentage, + + + : major percentage.
*Taken from literature (Forth et al., 1996; Mutschler, 1996; Hardman et al., 1996).
$Evaluated by using daily doses multiplied by the prescribed number of daily doses per year (Schwabe and Pa€rath, 1996).
%Conjugated only by phase II metabolism.
}Used literature: Abshagen et al., 1979.
}Used literature: Bendetta et al., 1995; Todd and Ward, 1988.
6Used literature: Landsdorp et al., 1990.
**Used literature: Frey and Janz, 1985.

range of lipid regulating agents and antiphlogistics (Stumpf et al., 1996a; Hirsch et al., 1996; Ternes et al.,
in STP e‚uents and river waters (Stumpf et al., 1998a; Ternes et al., 1998b).
1996a) as well as betablockers and b2-sympathomi- Firstly, a multi-method has been developed which
metics (Hirsch et al., 1996). Even estrogens have allows the determination of betablockers and b2-sym-
been detected in the aquatic environment by several pathomimetics as well as neutral drugs like diazepam or
carbamazepine at concentrations in the ng lÿ1-range after
authors in the ng lÿ1-range (Tabak and Bunch,
a joint solid phase extraction (Ternes et al., 1998a). For
1970; Rurainski et al., 1977; Tabak et al., 1981;
the detection of betablockers, b2-sympathomimetics by
Aherne and Briggs, 1989; Shore et al., 1993; Stumpf GC/MS a two step derivatization with silylation and
et al., 1996b). acetylation has been developed, whereas all other neutral
However, there is still a lack of information drugs can be measured without any derivatization.
about the behavior of drugs during their passage Alternatively, a time-saving method has been developed
through a STP and about the contamination of the using detection by LC-electrospray/MS/MS, in order to
aquatic environment due to the multitude of drugs determine neutral drugs (e.g. carbamazepine, phenazone,
applied in medicine. For instance in Germany ifosfamide and cyclophosphamide) down to 10 ng lÿ1 in
about 2900 di€erent drugs are allowed to be used in river water as well as in STP e‚uents (Ternes et al.,
human medicine (Rote Liste, 1994). The aim of this 1998a). The LC/MS/MS method was used for sewage,
work was to survey the exposure of German STP STP e‚uents and highly organic polluted river and stream
e‚uents and German rivers to drugs and some water samples. Secondly, a multi-method allows the deter-
mination of acidic drugs having a carboxylic moiety and
selected metabolites.
additionally in some cases one or two hydroxy groups
down to lower ng lÿ1-range (Ternes et al., 1998b). Lipid
regulating agents, antiphlogistics and three metabolites
EXPERIMENTAL SECTION with carboxylic and hydroxy groups are included within
this method. A separate determination of the acidic drugs
Methods
is also performable (Stumpf et al., 1996a) as for the
Several methods have been used for the determination metabolites. The acidic drugs and the ®ve metabolites
of drugs and their metabolites belonging to distinct medic-
were quanti®ed and con®rmed by GC/MS or GC/MS/MS
inal groups (see appendix) in the lower ng lÿ1-range includ-
ing solid phase extraction, derivatization, detection and after solid phase extraction and methylation by diazo-
con®rmation by GC/MS and GC/MS/MS or LC-electro- methane. All analysis of environmental samples have been
spray/MS/MS. The methods have already been published carried out by utilizing at least 5 point calibrations over
or submitted and are therefore only brie¯y mentioned the whole procedure of the previously described methods.

Table 2. Flow rates of a municipal sewage treatment plant near Frankfurt/Main at di€erent sampling periods
24.5±30.5.96 24.6±30.6.96 28.1±2.2.97 24.3±30.3.97 23.11±30.11.97

Average ¯ow rate (m3 dÿ1) 64,7202 12,500 54,300 26,430 55,3202 1,600 63,190 210,880 53,3002 6,200
lipid regulating lipid regulating
agents, betablockers, b2- agents, neutral drugs, ASA-
Investigated drugs antiphlogistics sympathomimetics antiphlogistics metabolites neutral drugs
Occurrence of drugs in sewage 3247

Sampling procedure Daily composite samples were taken from the rivers
Sampling of corresponding in¯uents and e‚uents of a Lahn (Oberbiel), Kinzig (Hanau), Fulda (Wahnhausen),
Werra (Heldra), Main (Bischofsheim), Rhine (Mainz),
STP. In ®ve sampling periods (24.5±30.5.96, 24.6±30.6.96,
Nidda (Nied) as well as from the stream Schwarzbach
28.1±2.2.97, 24.3±30.3.97 and 23.11±30.11.97) composite
(Trebur). Additionally, random samples were taken from
samples from a German municipal STP were taken daily
the Ruhr (Essen), Mosel (Wehlen), Neckar (Heidelberg),
from the raw in¯uent and the corresponding ®nal e‚uent
Elbe (Hamburg) and from 24 small rivers and streams
over a period of 6 d. Sampling was carried out by a ¯ow
mostly located in the Hessian Ried area (center of
proportional automatic sampler, whereby the composite Germany). From the river Rhine also daily composite
samples of the ®nal e‚uent were taken time related to the samples were taken over a 14 d sampling period (from
in¯uent. The municipal STP near Frankfurt/Main is con- January 1996 to June 1996 and September 1996 to
nected to about 312,000 population equivalents. It consists December 1996) at three locations situated at the Theodor
of three commonly used main treatment steps: preliminary Heuss bridge connecting Mainz and Wiesbaden.
clari®cation followed by an aerator tank with the addition Additionally, in December 1996 random samples were
of Fe(II)chloride for phosphate elimination and ®nally end taken from the river Main at di€erent distances between
point clari®cation. The average ¯ow rates at the sampling km 77 and km 4.5 from its entry point into the river
dates are listed in Table 2. All cooled water samples were Rhine.
analyzed within two days in order to keep microbial
degradation to a minimum.
Sampling of STP e‚uents and rivers. E‚uents of up to RESULTS AND DISCUSSION
49 municipal STP, treating mainly household discharges,
were sampled in order to analyze drug residues. All STP Elimination of drugs after passage through a munici-
consisted of three commonly used main treatment steps: pal STP
preliminary and ®nal clari®cation and an aerator tank.
Additionally, 43 STP treatment plants are equipped with The removal of drug residues during passage
phosphate elimination, 25 STP with a nitri®cation and 13 through a municipal STP near Frankfurt/Main was
with a denitri®cation treatment step. In general, random investigated by the analysis of corresponding daily
samples of the STP e‚uents were taken. The sampling
periods for the determination of acidic drugs like lipid reg- composite water samples of the raw in¯uent and
ulating agents and some antiphlogistics from STP e‚u- ®nal e‚uent after ®ltration (0.45 mm) at di€erent
ents, rivers and streams were from November 1995 to periods.
June 1996 and for the analysis of betablockers and b2- Average loads up to 3.0 kg dÿ1 (salicylic acid)
sympathomimetics from April 1996 to December 1996. were determined in the in¯uent and up to 114 g dÿ1
Neutral drugs like carbamazepine, phenazone or diazepam
and the metabolites of acetylsalicylic acid were screened in (carbamazepine) in the ®nal e‚uent. In Fig. 1 all
STP e‚uents as well as in rivers and streams from Octo- detected drugs are illustrated, which were mainly
ber 1996 to November 1997. present in the STP discharge. The elimination rates

Fig. 1. Elimination of di€erent drugs during passage through a municipal sewage treatment plant
located near Frankfurt/Main over 6 d. Sampling periods, betablockers, b2-sympathomimetics 24.6.±
30.6.97; antiphlogistics, lipid regulating agents: 28.1.±2.2.97; 24.5.±30.5.96; remaining drugs: 24.3.±
29.3.97; 23.11.±30.11.97.
3248 Thomas A. Ternes

of the investigated drugs during passage through some drugs like beza®brate, diclofenac, naproxen
the STP ranged from 7% (carbamazepine) to and clo®bric acid were signi®cantly reduced on the
greater than 99% (salicylic acid). Generally more rainfall day and did not recovered until the sixth
than 60% of the drug residues detected in the in¯u- day, with the exception being beza®brate. For beza-
ent were removed. Only carbamazepine, clo®bric ®brate a reduced elimination rate (<5%) was
acid, phenazone and dimethylaminophenazone observed even at the sixth day, the last day of the
showed lower average removal rates. Feno®brate, investigation period. These results indicate that the
acetaminophen and the metabolites of acetylsalicylic rainfall event may have been presumably respon-
acid (salicylic acid, o-hydroxyhippuric acid, gentisic sible for the decreased drug elimination rates by the
acid) were not detectable in the discharge of the STP. Whether these e€ects were caused by a
STP, although sometimes extremely high concen- reduced microbial activity or by altered sorption
trations up to 54 mg lÿ1 (salicylic acid) were deter- and ¯occulation conditions within this rain period
mined in the in¯uent. Hence, these drugs and or both, cannot be deduced. Presumably, the treat-
metabolites were eciently removed by the selected ment in the aerator tank with the simultaneously
municipal STP. It should be noted that the performed ¯occulation was mainly responsible for
observed elimination rates of drug residues during the observed drug removal rates. However, to prove
passage through the STP could not be di€erentiated this hypothesis further comparable investigations
in e€ects of sorption and biodegradation, since the concerning the sorption and biodegradation beha-
drugs sorption behavior on activated sludge was vior of drugs within the commonly utilized treat-
not investigated. However, most of the drugs exam- ment steps in municipal STP at di€erent seasons
ined in this study were incompletely eliminated and are essential to reveal the eciency of single treat-
thus contaminate the receiving water. ment processes.
The removal rates of several antiphlogistics and
lipid regulating agents was investigated during a Screening of drug levels in sewage plant e‚uents as
another sampling period (24.5.±30.5.96, Table 2), well as rivers and streams
which included a rainfall event on the fourth day Due to the incomplete elimination of many phar-
and led to an elevated ¯ow rate of about 50%, maceutical compounds and metabolites (e.g. clo®b-
from average 59,500 m3 dÿ1 to 89,900 m3 dÿ1 ric acid and feno®bric acid) in the selected
(Fig. 2). As shown in Fig. 2 the elimination rates of municipal STP near Frankfurt/Main a screening

Fig. 2. Elimination of drugs during passage through a municipal sewage treatment plant near
Frankfurt/Main over 6 d including a rainfall event: Sampling period, in¯uents: May 24th to May 29th
1996, e‚uents: May 25th to May 30th 1996.
Table 3. Concentrations of lipid regulators and metabolites in STP e‚uents as well as rivers and streams
STP e‚uents Rivers and streams
Substances in mg lÿ1 LOD* Number STP n>LOD* Median mg lÿ1 90-percentile Maximum mg lÿ1 LOD* Samples/rivers n>LOD* Median mg lÿ1 90-percentile Maximum mg lÿ1

Lipid regulating agents


Beza®brate 0.25 49 48 2.2 3.4 4.6 0.025 43/22 39 0.35 1.2 3.1
Gem®brozil 0.050 49 39 0.40 0.84 1.5 0.010 43/22 28 0.052 0.19 0.51
Feno®brate 0.050 20 2 n.d. n.d. 0.03 0.010 36/22 0 n.d. n.d. n.d.
Eto®brate 0.10 20 0 n.d. n.d. n.d. 0.030 36/22 0 n.d. n.d. n.d.
Clo®brate 0.10 20 0 n.d. n.d. n.d. 0.030 36/22 0 n.d. n.d. n.d.
Metabolites
Clo®bric acid 0.050 49 47 0.36 0.72 1.6 0.010 43/22 35 0.066 0.21 0.55
Feno®bric acid 0.050 49 41 0.38 0.68 1.2 0.010 43/22 26 0.045 0.17 0.28
*Limit of detection, n.d.: not detectable.

Table 4. Concentrations of antiphlogistics in STP e‚uents as well as rivers and streams


STP e‚uents Rivers and streams
Substances mg lÿ1 LOD* Number STP n>LOD* Median in mg lÿ1 90-percentile Maximum mg lÿ1 LOD*1 Samples/rivers n>LOD* Median in mg lÿ1 90-percentile Maximum mg lÿ1
Occurrence of drugs in sewage

Antiphlogistics
Diclofenac 0.050 49 49 0.81 1.6 2.1 0.010 43/22 43 0.15 0.80 1.20
Ibuprofen 0.050 49 42 0.37 1.2 3.4 0.010 43/22 35 0.07 0.28 0.53
Indometacine 0.050 49 49 0.27 0.40 0.60 0.010 43/22 35 0.04 0.17 0.20
Naproxen 0.050 10 10 0.30 0.42 0.52 0.010 20/20 20 0.070 0.15 0.39
Fenoprofen 0.050 49 0 n.d. n.d. n.d. 0.010 43/22 0 n.d. n.d. n.d.
Ketoprofen 0.050 49 37 0.20 0.25 0.38 0.010 43/22 5 n.d. 0.12 0.12
Phenazone 0.10 30 28 0.16 0.30 0.41 0.020 26/20 21 0.024 0.15 0.95
Acetaminophen 0.50 49 4 n.d. n.d. 6.0 0.150 31/16 0 n.d. n.d. n.d.
Acetylsalicylic acid 0.10 49 22 0.22 0.32 1.5 0.020 43/22 17 n.d. 0.16 0.34
Dimethylaminophenazone 0.10 16 3 n.d. 0.15 1.0 0.030 26/20 2 n.d. n.d. 0.34
Meclofenamic acid 0.050 10 0 n.d. n.d. n.d. 0.010 43/22 0 n.d. n.d. n.d.
Tolfenamic acid 0.050 10 0 n.d. n.d. n.d. 0.010 30/20 0 n.d. n.d. n.d.
Metabolites
Salicylic acid 0.050 36 9 n.d. 0.063 0.14 0.010 35/19 24 0.025 0.13 4.1
o-hydroxyhippuric acid 0.20 36 0 n.d. n.d. n.d. 0.075 35/19 0 n.d. n.d. n.d.
Gentisic acid 0.20 36 3 n.d. 0.20 0.59 0.075 35/19 5 n.d. 0.11 1.2
*Limit of detection, n.d.: not detectable.
3249
3250

Table 5. Concentrations of betablockers, b2-sympathomimetics in STP e‚uents as well as rivers and streams
STP e‚uents Rivers and streams
Substances mg lÿ1 LOD* Number STP n>LOD* Median in mg lÿ1 90-percentile Maximum mg lÿ1 LOD* Samples/rivers n>LOD* Median in mg lÿ1 90-percentile Maximum mg lÿ1

Betablockers
Metoprolol 0.025 29 29 0.73 1.3 2.2 0.010 45/23 38 0.045 1.2 2.2
Propranolol 0.025 29 28 0.17 0.23 0.29 0.010 45/23 26 0.012 0.44 0.59
Nadolol 0.025 29 20 0.025 0.042 0.06 0.010 45/23 0 n.d. n.d. n.d.
Carazolol 0.025 29 13 n.d. 0.070 0.12 0.010 45/23 3 n.d. 0.10 0.11
Timolol 0.025 29 2 n.d. n.d. 0.07 0.010 45/23 1 n.d. n.d. 0.01
Betaxolol 0.025 29 17 0.057 0.10 0.19 0.010 45/23 1 n.d. n.d. 0.028
Bisoprolol 0.025 29 17 0.057 0.13 0.37 0.010 45/23 19 n.d. 0.19 2.9
b2-sympathomimetics
Fenoterol 0.050 29 1 n.d. n.d. 0.060 0.010 45/23 1 n.d. n.d. 0.061
Terbutalin 0.050 29 11 n.d. 0.087 0.12 0.010 45/23 0 n.d. n.d. n.d.
Salbutamol 0.050 29 10 n.d. 0.072 0.17 0.010 45/23 2 n.d. n.d. 0.035
Clenbuterol 0.050 29 5 n.d. n.d. 0.08 0.010 45/23 2 n.d. n.d. 0.050
*Limit of detection, n.d.: not detectable.
Thomas A. Ternes

Table 6. Concentrations of diazepam, carbamazepine and anticancer agents in STP e‚uents as well as rivers and streams
STP e‚uents Rivers and streams
Samples/
Substances mg lÿ1 LOD* Number STP n>LOD* Median in mg lÿ1 90-percentile Maximum mg lÿ1 LOD* rivers n>LOD* Median in mg lÿ1 90-percentile maximum mg lÿ1

Psychiatric drugs
Diazepam 0.030 20 8 n.d. 0.03 0.04 0.030 30/20 0 n.d. n.d. n.d.
Antiepileptic drugs
Carbamazepine 0.050 30 30 2.1 3.7 6.3 0.030 26/20 24 0.25 0.82 1.1
Cytostatic agents
Ifosfamide 0.010 16 2 n.d. 0.040 2.9 0.010 26/20 0 n.d. n.d. n.d.
Cyclophosphamide 0.010 16 4 n.d. 0.018 0.020 0.010 26/20 0 n.d. n.d. n.d.
*Limit of detection, n.d.: not detectable.
Occurrence of drugs in sewage 3251

program of up to 49 di€erent German municipal tration for diclofenac was 0.81 mg lÿ1 in STP
STP e‚uents has been conducted. Finally, German e‚uents and 0.15 mg lÿ1 in river and stream waters,
rivers and streams were screened in order to assess whereas the median values of the other antiphlogis-
the occurrence of drug residues. The analytical data tics were lower by at least a factor of two. The anti-
are listed in Tables 3±6. phlogistics, present in the rivers and STP e‚uents,
Lipid regulating agents. The lipid regulators beza- were mainly applied in human medicine (Mutschler,
®brate and gem®brozil were present in the majority 1996; Forth et al., 1996). Only naproxen is ad-
of German municipal STP e‚uents investigated in ditionally used in veterinary medicine with appreci-
this study (Table 3). For beza®brate concentrations able amounts (LoÈscher et al., 1994). However, the
up to 4.6 mg lÿ1 and median values surprisingly as veterinary drugs meclofenamic acid and tolfenamic
high as 2.2 mg lÿ1 have been observed. Rivers and acid could not be detected at all. Therefore, it is
streams were also polluted by these drugs, however likely that the detected antiphlogistics and lipid reg-
median concentrations declined by a factor between ulators arise predominately from human application
5 and 10, as compared to STP discharges. underlined by the ubiquitous contamination of mu-
The unpolar lipid regulators (clo®brate, eto®brate nicipal STP discharges.
and feno®brate) were in general not detectable in In a few STP e‚uents, acetylsalicylic acid, aceta-
STP e‚uents, but their polar metabolites clo®bric minophen and dimethylaminophenazone were
acid and feno®bric acid were present with concen- detectable (Table 4). Their maximum concentrations
tration levels up to 1.6 mg lÿ1 and 1.2 mg lÿ1 respect- exceeded 1 mg lÿ1, but the 90-percentile values were
ively. Concentration levels within the ng lÿ1-range already not detectable or dramatically decreased.
were observed for the above two metabolites in riv- However, acetylsalicylic acid was detected in 22 of
ers and streams. These two examples illustrate the 49 STP e‚uents at concentrations up to 1.5 mg lÿ1
importance of the principal metabolites for the con- and in rivers and streams up to 0.34 mg lÿ1,
tamination of the aquatic environment, due to their although it is ultimately biodegradable in labora-
frequently increased polarity. Clo®bric acid is the tory test systems as reported by Richardson and
active metabolite of the lipid regulators clo®brate,
Bowron (1985). Acetaminophen was not present in
etofyllinclo®brate and eto®brate, whereas feno®bric
surface waters above the detection limits, presum-
acid is only derived from feno®brate (Balfour et al.,
ably due to high removal eciencies of municipal
1990; Forth et al., 1996; Mutschler, 1996). The
STP. Above 98% have been eliminated in the inves-
complete hydrolysis of the original drugs to feno®b-
tigated STP near Frankfurt/Main.
ric acid and clo®bric acid occurs immediately after
Dimethylaminophenazone was detected in two
intake. Their principal excretion products are glu-
samples of the river Main at concentrations up to
curonides of the acidic metabolites, which are
0.34 mg lÿ1. Its approval as human drug in the
formed by conjugation of the carboxylic moiety
European Union was withdrawn, because when it
with glucuronic acid. The percentage of the glucur-
comes in contact with nitrite at an acidic pH, the
onides excreted exceeded mostly 60% (Table 1),
whereas the portion of the non-conjugated acids are carcinogenic dimethylnitrosamine is formed (Forth
generally very small (below 10%). Original non- et al., 1996). The presence of dimethylaminophena-
hydrolyzed drugs are not excreted, but other metab- zone in municipal STP discharges may be caused by
olites like reduced feno®bric acid (Balfour et al., its low application level in veterinary medicine
1990) are present in urine or faeces in small (LoÈscher et al., 1994), but in the river Main the
amounts as glucuronide or unmetabolized. Because concentrations are more likely derived by industrial
the concentration levels of clo®bric acid and feno- discharges.
®bric acid were as high as 1.6 mg lÿ1 in STP e‚u- Gentisic acid, a metabolite of acetylsalicylic acid
ents, it appears most likely that their glucuronide- as well as o-hydroxyhippuric acid and salicylic acid
conjugates were at least partially cleaved in sewage (Mutschler, 1996; Forth et al., 1996) were not, or
or during passage through a municipal STP and only sporadically found in municipal STP e‚uents.
thus increase the relevant environmental concen- In the investigated STP near Frankfurt 98% elimin-
trations. ation rate was observed. However, salicylic acid
Antiphlogistics. 6 of the 12 selected antiphlogistics was only detected in 33 of 49 samples taken from
(analgesic-antipyretic and antiin¯ammatory agents) river and stream waters, with a median concen-
were detected in more than 50% of the investigated trations of 0.2 mg lÿ1. In addition to the (bio)degra-
municipal STP e‚uents as listed in Table 4. Mainly dation of acetylsalicylic acid other reasons for its
diclofenac, ibuprofen, indometacine, naproxen, observed concentrations may be a result of an ap-
ketoprofen and phenazone were present in the STP plication as an antioxidant agent, industrial STP
discharges. The 90-percentile concentrations of discharges or natural occurrence may be possible to
diclofenac and ibuprofen exceeded 1 mg lÿ1. With cause the observed concentrations. For example,
the exception of ketoprofen these drugs were also salicylic acid was formed by sulfate-reducing bac-
detected in rivers and streams, with diclofenac hav- teria obtained from a shallow anoxic aquifer by o-
ing the highest concentrations. The median concen- cresol degradation (Su¯ita et al., 1989).
3252 Thomas A. Ternes

Due to the high application level of antiphlogis- tration of 0.088 mg lÿ1 and cyclophosphamide with
tics indicated by high annual prescriptions (Table 1), 0.019 mg lÿ1. Ifosfamide was present in the e‚uent
and the opportunity to purchase most of them of one additional rural municipal STP (connected
without prescription, their occurrence in the aquatic to about 15,000 population equivalents), but sur-
environment is plausible. prisingly as high as 2.9 mg lÿ1. Depending on the
Betablockers and b2-sympathomimetics. The beta- origin of the sewage exceptionally ifosfamide and
blockers metoprolol and propranolol, as well as cyclophosphamide can obviously be present in mu-
betaxolol, bisoprolol and nadolol in smaller con- nicipal STP discharges. However, a wide-spread
centrations were detectable in STP discharges, and presence was not observed and appears unlikely,
with the exception of nadolol additionally in rivers because their application is mostly con®ned to che-
and streams (Table 5). However, only for metopro- motherapy in hospitals. Steger-Hartmann et al.
lol and propranolol median values in river and (1996) found ifosfamide and cyclophosphamide in
streams have been observed above the detection
the sewage of a university hospital with concen-
limits. The highest median values were found for
trations of 0.024 mg lÿ1 and 0.146 mg lÿ1 respectively.
metoprolol, with a concentration of 0.73 mg lÿ1 in
They assumed that these anticancer agents are not
STP e‚uents and 0.45 mg lÿ1 in the river and
degraded during passage through a STP, because a
stream waters. The 90-percentile values of both
signi®cant reduction did not occur in their labora-
matrices (above 1 mg lÿ1) revealed that high concen-
trations of metoprolol were occasionally present in tory-scale sewage treatment plant operated accord-
the aquatic environment, although it is mainly ing to the modi®ed OECD instruction 303A.
excreted as a metabolite. Obviously, the excretion Carbamazepine. The antiepileptic drug carbama-
rates of the unmetabolized drugs are suciently zepine was ubiquitously present in the aquatic en-
high to cause the observed environmental concen- vironment. Carbamazepine showed a relatively high
trations, because other application areas than that median value of 2.1 mg lÿ1 in German STP e‚uents
of medicine are not reported in literature. However, and 0.25 mg lÿ1 in German river waters (Table 6).
their principal excreted metabolites (e.g. metoprolol The highest maximum concentration detected for
acid, 4-hydroxypropanolol, a-naphtoxylactic acid) all drugs investigated in STP e‚uents was detected
are probably present in comparable concentrations for carbamazepine at a concentration of 6.3 mg lÿ1
(BalmeÂr et al., 1987; Walle et al., 1988), but unfor- and a 90-percentile value as high as 3.7 mg lÿ1.
tunately they are not commercially available as Annual prescriptions of carbamazepine in Germany
reference compounds and must therefore be syn- are as high as approximately 80 tons per year
thesized prior to analysis. (Table 1). However, as revealed by pharmacokineti-
The b2-sympathomimetics terbutalin and salbuta- cal data only 1±2% of carbamazepine is excreted
mol were sometimes detectable in STP e‚uents, unmetabolized. The major metabolite in humans is
whereas clenbuterol and fenoterol were only 10, 11 epoxy-carbamazepine, which is hydrolyzed
detected in few cases. However, the concentrations further to diol-derivatives and are excreted princi-
of all b2-sympathomimetics were extremely low and pally as glucuronides (Frey and Janz, 1985; Hard-
did not exceed 0.2 mg lÿ1. In river and stream waters man et al., 1996). But additionally carbamazepine
they were only sporadically present above the is inactivated by hydroxylation of the aromatic ring
detection limit of 0.01 mg lÿ1, with a maximum con- or N-glucuronidation at the carbamoyl moiety.
centration of 0.061 mg lÿ1. These glucuronide-conjugates can presumably be
Diazepam and two antineoplastics. Diazepam, a cleaved in sewage and STP and thus increase the
psychiatric drug, was present in 8 of 20 STP dis-
environmental concentrations. On the other hand
charges at relative low concentrations amounting
the longer the medical treatment with carbamaze-
up to 0.04 mg lÿ1. In river and stream waters diaze-
pine the higher the eciency of metabolization,
pam was not detected above the determination
because of auto-induction of the drug metabolizing
limit of 0.030 mg lÿ1 (Table 6).
enzymes (e.g. microsomale mono-oxygenase
The frequently in cancer chemotherapy applied
antineoplastics ifosfamide and cyclophosphamide enzymes). If occasionally industrial discharges or
(Mutschler, 1996; Forth et al., 1996) were detected even other application areas may increase the en-
in few municipal STP e‚uents, but not in the inves- vironmental concentrations, can yet not be con-
tigated rivers and streams (Table 6). All samples cluded. However, the principal reason for the
were analyzed by the method containing LC-elec- ubiquitously high carbamazepine concentrations
trospray/MS/MS. Detection limits were 0.010 mg lÿ1 appears to be the extremely low removal rate in
for both matrices. Cyclophosphamide was only pre- municipal STP. Only 7% were removed in the
sent in STP discharges at concentrations up to selected STP near Frankfurt (Fig. 1). Although in
0.02 mg lÿ1, whereas ifosfamide was observed in two the raw in¯uent several drugs were present in com-
samples above 0.08 mg lÿ1. In the discharge of a parable or even higher loads, in the discharge of
STP, which is associated to a large university the selected STP carbamazepine was found to be
hospital, ifosfamide was detected with a concen- present with the highest average load (114 g dÿ1).
Occurrence of drugs in sewage 3253

Fig. 3. Scheme for the main fates of drugs in the environment after application (STP: sewage treatment
plant).

Fate of drugs in the environment from human applications, indicated by their wide-
The detected drugs residues in German rivers spread distribution in municipal STP. The fate of
and streams (Tables 3±6) presumably arise mainly veterinary and human drugs after urinal or faecal

Fig. 4. Comparative contamination of small streams located in Hessisches Ried/Germany and the large
rivers Rhine (Mainz) and Main (Bischofsheim) by selected acidic drugs and betablockers (Mean values
of 14 d composite values of the Rhine (Mainz) from Jan. to Dec. 1996; mean values for random
samples of the river Main (Bischofsheim) from Dec. 1996; mean values for random samples of streams
at three sampling periods (April, September, October) in 1996.
3254 Thomas A. Ternes

Fig. 5. Contamination of the river Rhine (Mainz) by acidic drugs in 1996 (14 daily composite samples
taken at Mainz between January 1996 to June 1996 and September 1996 to December 1996).

excretion are quite di€erent to each other. In gen- and ground water may be caused by the application
eral German municipal sewage and therefore of digested sludge from municipal STP on agricul-
excreted human pharmaceuticals have to pass tural areas. Additionally, transport of drugs via
through a STP prior to entering rivers or streams, bank ®ltration from highly contaminated surface
whereas veterinary drugs are more likely to con- water into ground water is also a possibility, as the
taminate soil and ground water (without previous in®ltration of waste waters directly from leakages
waste water treatment) when liquid manure is used
in drains (Hall and Medmenham, 1992). Also drugs
for top soil dressing (Fig. 3). Additionally, after
disposed together with domestic waste can reach
rainfall incidents surface waters can be polluted
land®ll sites which could lead to ground water con-
with human or veterinary drugs by run-o€ from
®elds treated with digested sludge or livestock slur- tamination by leaching (Holm et al., 1995). A
ries respectively. Industrial waste water may be unique pathway for the contamination of soil and
another possible source for the contamination of ground water by medicinal residues, derived from
surface waters (Richardson and Bowron, 1985), but human application, may be the disposal of raw
are surely not responsible for their ubiquitous sewage or STP e‚uents by spray and broad irriga-
occurrence. Another possible contamination of soil tion in agricultural areas.

Table 7. Acidic drugs in the river Main (sampling date: Dec. 1996) at di€erent distances between 77 and 4.5 km from its entry point into
the river Rhine
77 km 56 km 46 km 34 km 26 km 16 km 4.5 km
(conc. in mg lÿ1) LOD* 12.12.96 12.12.96 13.12.96 13.12.96 14.12.96 14.12.96 15.12.96

Acetylsalicylic
acid 0.020 n.d. n.d. n.d. n.d. n.d. n.d. n.d.
Clo®bric acid 0.020 n.d. n.d. n.d. 0.02 n.d. 0.02 0.03
Ibuprofen 0.020 n.d. n.d. n.d. 0.02 n.d. n.d. 0.02
Gem®brozil 0.020 n.d. n.d. 0.02 0.03 n.d. 0.02 0.02
Fenoprofen 0.020 n.d. n.d. n.d. n.d. n.d. n.d. n.d.
Ketoprofen 0.020 n.d. n.d. n.d. n.d. n.d. n.d. n.d.
Diclofenac 0.020 0.09 0.07 0.10 0.14 0.09 0.10 0.14
Feno®bric acid 0.020 n.d. n.d. n.d. n.d. n.d. n.d. 0.03
Beza®brate 0.10 0.10 n.d. 0.10 0.15 0.09 0.10 0.20
Indometacine 0.020 n.d. 0.02 0.03 0.03 n.d. 0.03 0.02
*Limit of detection, n.d.: not detectable.
Occurrence of drugs in sewage 3255

Comparison of drug residues in large rivers and aquatic environment sometimes in comparable or
streams even higher concentrations to the original
Acidic drugs as well as the betablockers metopro- (unchanged) drugs indicated by the environmental
lol and propranolol were found in small rivers and concentrations of clo®bric acid and feno®bric acid.
streams, as illustrated in Fig. 4, at higher concen- For the majority of the medicinal products includ-
trations as in the larger rivers, Rhine and Main. In ing their metabolites an environmental risk assess-
general residues of acidic drugs in the river Rhine ment had never been carried out, although they are
(Mainz) and Main (Bischofsheim) ranged from applied in the ton-range per year in medicine and
0.05 mg lÿ1 to 0.3 mg lÿ1 (Fig. 5, Table 7). With the are present in the environment presumably for
exception of phenazone and dimethylaminophena- many years.
zone the maximum and 90-percentile concentrations The chemotherapeutic alkylating agents ifosfa-
of all drugs listed in Tables 3±6 were always mide and cyclophosphamide may presumably be
observed in streams and not in rivers. Phenazone attributed to adverse environmental e€ects, due to
and dimethylaminophenazone were present in the their enormous side e€ects and their unspeci®c
river Main up to 0.95 mg lÿ1, presumably mainly alkylating properties (Hardman et al., 1996;
due to an industrial STP discharge. Additionally, a Mutschler, 1996), but fortunately they were not
multitude of drugs (up to 20) were detectable in detectable in rivers and streams above the detection
several investigated streams up to concentration limit of 10 ng lÿ1. However, exceptions may occur
greater than 6 mg lÿ1. These streams, located in the as indicated by few contaminations of municipal
Hessian Ried (center of Germany) contain a high STP discharges (Table 6).
percentage of STP discharges, as indicated by their Many drugs interact with human receptors (e.g.
relatively high boron concentrations, ranging from betablockers: antagonists of b-adrenergic receptors)
0.25 to 1.0 mg lÿ1 (Fooken et al., 1997). Due to the when causing the required pharmacological e€ects.
fact that the STP discharges are highly contami- Aquatic vertebrates frequently have similar recep-
nated by those drugs which are present in these tors which may be favorable for these drugs.
streams, it can be assumed, that the enhanced per- However, the amounts of drugs in rivers and es-
centage of municipal STP e‚uents are responsible pecially streams are several magnitudes lower com-
for the elevated concentrations of the drugs (e.g. pared to those applied in medicine; but it can not
lipid regulating agents, betablockers). The run-o€ be ruled out that the multitude of drugs, which are
from agricultural areas, another potential source present in rivers and streams, have adverse e€ects
for the intake of drugs, was negligible in this region on aquatic organisms. Nevertheless, by these low
(Hessian Ried). Firstly, the amounts of digested environmental drug concentrations rather chronical
municipal sludge applied to this area was very low e€ects may be caused than acute toxic e€ects.
and secondly run-o€ events are extremely seldom,
due to the excellent permeability of the soils in CONCLUSIONS
large areas in this region (Berthold et al., 1993).
The occurrence of a multitude of di€erent drugs
in German rivers and streams indicates the rela-
ENVIRONMENTAL RISK ASSESSMENT
tively high stability of these medicinal compounds
under environmental conditions. Obviously, the
An environmental risk assessment for the drugs common sewage treatment process is insucient to
detected in river and stream waters was not under- completely eliminate these drug residues.
taken, due to the lack of available ecotoxicological Additionally, metabolites formed by conjugation
data (RoÈmbke et al., 1996). In the European Union (e.g. with glucuronic acid, sulfate) within phase II
physico±chemical data like Koc or Pow and in par- reactions are likely to be cleaved in the environ-
ticular ecotoxicological data like LC50, EC50 and ment into the original (unchanged) pharmaceuticals
NOEC are not mandatory for the governmental (Richardson and Bowron, 1985), and hence may
approval of human medicinal products. However, increase the relevant environmental concentrations.
since January 1995 for the use of veterinary drugs In Germany approximately 2900 drugs have
in Germany an environmental risk assessment may approval for the use in human medicine (Rote
be required before the approval is granted by the Liste, 1994). The investigated 32 drugs and 5
government, when signi®cant exposure to drugs is metabolites represents only about 1% of the
expected and when the drug is potentially harmful approved pharmaceutical compounds when consid-
to the environment. Additionally, pharmacokineti- ering human and veterinary drugs. However, the
cal studies have indicated that drugs are mainly detected drugs were screened on the basis of their
excreted as metabolites. Due to their enhanced po- high prescriptions in human medicine, but it can be
larity, it can be assumed that the metabolites were assumed that many other drugs and especially
often poorly eliminated during passage through a polar metabolites were present in the aquatic en-
water treatment process of a STP. Hence, drug vironment at concentrations up to the mg lÿ1-range.
metabolites are expected to be present in the Investigations about the contamination of di€erent
3256 Thomas A. Ternes

ground water types by drug residues are currently tural wastes, eds., J. E. Hall, P. L. `Hermite and P. J.
under way. Newmann, Review of COST 68/681 programme, 1972±
1990.
Heberer T. and Stan H.-J. (1996) Vorkommen von pola-
AcknowledgementsÐThis study was ®nancially supported ren organischen Kontaminanten im Berliner
by grants obtained from the Ministry of Education and Trinkwasser. Occurrence of polar organic contaminants
Research (BMBF)/Germany and the Hessian Ministry for in Berlin drinking water. Vom Wasser 86, 19±31.
Environment, Energy, Youth, Family and Health. I also Hignite C. and Azarno€ D. L. (1977) Drugs and drug
wish to express my thanks to Professor Haberer and metabolites as environmental contaminants:
Professor Wilken for supporting this work and for their Chlorophenoxyisobutyrate and salicylic acid in sewage
useful discussions. In particular I thank Roman Hirsch water e‚uent. Life Sci. 20, 337±342.
(Dipl.-Chem.), Marcus Stumpf (Dipl.-Ing. (TH)), BaÈrbel Hirsch R., Ternes T., Haberer K. and Kratz K.-L. (1996)
Schuppert (Dipl.-Ing. (FH)) and Jutta MuÈller (Dipl.-Ing. Nachweis von Betablockern und Bronchospasmolytika
(FH)) for their accurately performed laboratory works in der aquatischen Umwelt. Determination of beta-
and Peter Seel (Dipl.-Chem. Dipl.-Biol.) of the Hessische blocker and b2-sympatomimetics in the aquatic environ-
Landesanstalt fuÈr Umwelt (HLfU) for providing samples ment. Vom Wasser 87, 263±274.
of the Hessian rivers, local streams and Hessian STP e‚u- Holm J. H., RuÈgge K., Bjerg P. L. and Christensen T.
ents. H. (1995) Occurrence and distribution of pharmaceuti-
cal organic compounds in the ground water downgradi-
ent of a land®ll (Grinsted, Denmark) . Environ. Sci.
REFERENCES Technol. 29, 1415±1420.
Landsdorp D., Vree T. B., Janssen T. J. and Guelen P.
Abshagen U., Bablok W., Koch K., Schmidt H. A. E., J. (1990) Pharmacokinetics of retal Dilclofenac and its
Senn M. and Stork H. (1979) Disposition pharmacoki- hydroxy metabolites in man. Int. J. Clin. Pharmacol.
netics of beza®brate in man. European J. Clin. Ther. Toxicol. 28, 298±302.
Pharmacol. 16, 31±38. LoÈscher W., Ungemach F. R. and Kroker R. (1994)
Aherne G. W. and Briggs R. (1989) The relevance of the Grundlagen der Pharmakotherapie bei Haus und
presence of certain synthetic steroids in the aquatic en- Nutztieren. Basics of pharmacological therapy of dom-
vironment. J. Pharm. Pharmacol. 41, 735±736. estic animals. Verlag Paul Parey, Berlin, Hamburg.
Balfour J. A., McTarish D. and Heel R. C. (1990) A Mutschler E. (1996) Arzneimittelwirkungen, Lehrbuch der
review of its pharmacodynamic and pharmacokinetic Pharmakologie und Toxikologie, Textbook of pharma-
propertiesand therapeutic use in dyslipidaemia. Drugs cology and toxicology. Wissenschaftliche
40, 260±290. Verlagsgesellschaft mbH, Stuttgart.
BalmeÂr K., Zhang Y., LagerstroÈm P.-O. and Persson B.- Richardson M. L. and Bowron J. M. (1985) The fate of
A. (1987) Determination of metoprolol and two major pharmaceutical chemicals in the aquatic environment. J.
metabolites in plasma and urine by column liquid chro- Pharm. Pharmacol. 37, 1±12.
matography and ¯uorometric detection. J. Chrom. 417, Rogers I. H., Birtwell I. K. and Kruzynski G. M. (1986)
357±365. Organic extractables in municipal wastewater
Bendetta C., Sallustio B. and Fairchild B. A. (1995) Vancouver, British Columia. Water Poll. Res. J. Can.
Biosynthesis, characterisation and direct high-perform- 21, 187±204.
ance liquid chormatographic analysis of gem®brozil 1-o- RoÈmbke J., Knacker Th. and Stahlschmidt-Allner P.
b-acyl-glucuronide. J. Chrom. 665, 345±353. (1996) Studie uÈber Umweltprobleme im Zusammenhang
Berthold G., Grieb H., Konarski C. U., LoÈns-Hana C.,
mit Arzneimitteln. Study about enviromental problems in
Meyer U., Von Pape W. P. and Toussaint B. (1993)
context with drugs. F + E Vorhabens Nr. 106 04 121,
Hydrogeologie in Hessen Handbuch Teil 3,
Umweltbundesamt, Berlin.
Grundwasserbescha€enheitsbericht 1993 Hessen.
Rote Liste (1994) Arzneimittelverzeichnis des BPI. Drug
Hydrogeology in Hessian, Handbook, Part 3,
register of BPI, ed., Bundesverband der pharmazeu-
Conditions of Ground Water Quality.
tischen Industrie e.V., ECV Editor Cantor Verlag fuÈr
Fooken C., Gihr R., HaÈckle M. and Seel P (1997)
Orientierende Messungen gefaÈhrlicher Sto€e ± Medizin und Naturwissenschaften GmbH.
Landesweite Untersuchungen auf organische Rurainski R. D., Theiss H. J. and Zimmermann W. (1977)
Spurenverunreiningungen in hessischern FlieûgewaÈssern, UÈber das Vorkommen von natuÈrlichen und synthe-
AbwaÈssern und KlaÈrschlamm 1991±1996. Orientated tischen OÈstrogenen im Trinkwasser. Occurrence of natu-
measurements of polluting compounds ± nationwide inves- ral and synthetic estrogenes in drinking water. Gwf-
tigations for organic trace pollutants in Hessian running Wasser/Abwasser 118, 287±291.
waters, sewages and activated sludge 1991±1996. Schwabe U. and Pa€rath D. (1995) Arzneiverord-
Umweltplanung, Arbeits- und Umweltschutz, 233 nungsreport Â95 Aktuelle Daten, Kosten, Trends und
Schriftenreihe der Hessischen Landesanstalt fuÈr Kommentare. Drug prescribing report `95 Current data,
Umwelt, Wiesbaden. costs, trends and comments, Gustav Fischer Verlag,
Forth W., Henschler D., Rummel W. and Starke K. Stuttgart, Jena.
(1996) Allgemeine und spezielle Pharmakologie und Schwabe U. and Pa€rath D. (1996) Arzneiverord-
Toxikologie. Wissenschaftsverlag Mannheim/Leipzig/ nungsreport Â96 Aktuelle Daten, Kosten, Trends und
Wien/ ZuÈrich, 6. Au¯age. Kommentare. Drug prescribing report `96 Current data,
Frey H.-H and Janz D. (1985) Antiepileptic Drugs, costs, trends and comments, Gustav Fischer Verlag,
Handbook of Experimental Pharmacology, Vol. 74. Stuttgart, Jena.
Springer Verlag Berlin, Heidelberg, New York, Tokyo. Shore L. S., Gurevitz M. and Shemesh M. (1993)
Hardman J. G., Limbird L. E., Molino€ P. B., Ruddon Estrogen as an environmental pollutant. Bull. Environ.
R. W. and Gilman, A. G. (1996) Goodman and Gilman's Contam. Toxicol. 51, 361±366.
The Pharmacological Basis of Therapeutics. MacGraw- Stan H.-J. and LinkerhaÈger M. (1994) Vorkommen von
Hill Health Professions Divisions, New York. Clo®brinsaÈure im aquatischen System -FuÈhrt die thera-
Hall J. E. and Medmenham W. (1992) Agricultural value peutische Anwendung zu einer Belastung von
of organic sludge and liquid agricultural waters In Ober¯aÈchen-, Grund- und Trinkwasser. Occurrence of
Treatment and use of sewage sludge and liquid agricul- clo®bric acid in aquatic systems ± Does the medical ap-
Occurrence of drugs in sewage 3257

plication cause a contamination of surface, ground and as water pollutants I. Dev. Ind. Microbiol. 11,
drinking water. Vom Wasser 83, 57±68. 367±376.
Steger-Hartmann Th., KuÈmmerer K. and Schecker Ternes T. A., Hirsch R., MuÈller J. and Haberer K. (1998)
J. (1996) trace analysis of the antineoplastics ifosfamide Method for the determination of neutral drugs as well
and cyclophosphamide in ewage water by two-step as b2-sympathomimetics in aqueous matrices using GC/
solid-phase extraction and gas chromatography-mass MS and LC/MS/MS. Fres. J. Anal. Chem. (accepted).
spectrometry. J. Chrom. 726, 179±184. Ternes T. A., Stumpf M., Schuppert B. and Haberer K.
Stumpf M., Ternes T. A., Haberer K., Seel P. and (1998) Determination of acidic drugs and antiseptics in
Baumann W. (1996a) Nachweis von Arzneimit- sewage and river water. Vom Wasser (in press).
telruÈckstaÈnden in KlaÈranlagen und FlieûgewaÈssern. Todd P. A. and Ward A. (1988) A review of ist pharma-
Determination of drugs in sewage treatment plants and codynamic and pharmacokinetic properties, and thera-
river water. Vom Wasser 86, 291±303. peutic use in dyslipidaemia. Drugs 36, 314±339.
Stumpf M., Ternes T., Haberer K. and Baumann Waggott A. (1981) Trace Organic Substances in the River
W. (1996b) Nachweis von natuÈrlichen OÈstrogenen in Lee [Great Britain].Chem. Water Reuse, ed., W. J.
KlaÈranlagen und FlieûgewaÈssern. Determination of Cooper, Vol. 2, pp. 55±99. Ann. Arbor Sci., Ann
natural estrogens in sewage treatement plants and river Arbor, MI.
water. Vom Wasser 87, 251±261. Walle Th., Webb J. G., Bagwell E. E., Walle U. K.,
Su¯ita J. M., Liang L. and Saxena A. (1989) The anaero- Daniell H. B. and Ga€ney Th. E. (1988) Stereoselective
bic biodegradation of o-, m- and p-cresol by sulfate- delivery and actions of beta receptor antagonists.
reducing bacterial enrichment cultures obtained from a Biochem. Pharmacol. 37, 115±124.
shallow anoxic aquifer. J. Industrial Microbiol. 4, 255± Watts C. D., Crathorne M., Fielding M. and Steel C. P.
266. (1983) Identi®cation of non-volatile organics in water
Tabak H., Bloomhu€ R. N. and Bunch R. L. (1981) using ®eld desorption mass spectrometry and high per-
Steroid hormones as water pollutants II. Dev. Ind. formance liquid chromatography. In Analysis of organic
Microbiol. 22, 497±519. micropollutants in water, ed., G. Angeletti et al. pp.
Tabak H. and Bunch H. L. (1970) Steroid hormones 120±131. Reidel Publ. Corp., Dordrecht.

APPENDIX OVERLEAF
3258 Thomas A. Ternes

APPENDIX

Name, CAS-registration Number, Chemical Structures And Medicinal Classes Of The Analyzed Drugs
A1: Antiphlogistics, lipid regulating agents and corresponding metabolites
Occurrence of drugs in sewage 3259

A2: Betablockers and b2-sympathomimetics

Ðcontinued overleaf
3260 Thomas A. Ternes

A3: Metabolites of acetylsalicylic acid and neutral drugs of di€erent medicinal classes