Environment International 39 (2012) 73–86

Contents lists available at SciVerse ScienceDirect

Environment International
journal homepage: www.elsevier.com/locate/envint

Anticancer drugs in surface waters

What can we say about the occurrence and environmental significance of cytotoxic,
cytostatic and endocrine therapy drugs?
Jean-Philippe Besse a,⁎, Jean-François Latour b, 1, Jeanne Garric a, 2
a
Cemagref, UR Milieux Aquatiques Ecologie et Pollution (MAEP), Laboratoire d'écotoxicologie/Laboratoire d'analyses physico-chimiques des milieux aquatiques, 3 bis quai Chauveau,
CP 220, F-69226 Lyon, France
b
Pharmacie, Centre Léon Bérard, 28 rue Laennec 69008 Lyon, France

a r t i c l e i n f o a b s t r a c t

Article history: This study considers the implications and research needs arising from anticancer (also referred to as antineo-
Received 20 June 2011 plastic) drugs being released into the aquatic environment, for the entire therapeutic classes used: cytotoxic,
Accepted 6 October 2011 cytostatic and endocrine therapy drugs.
Available online 8 November 2011
A categorization approach, based on French consumption amounts, allowed to highlight parent molecules
and several metabolites on which further occurrence and ecotoxicological studies should be conducted.
Keywords:
Anticancer drugs
Investigations of consumption trends at a national and a local scale show an increase in the use of anticancer
Antineoplastic drugs between 2004 and 2008, thus leading to increased levels released in the environment. It therefore
Pharmaceuticals appears necessary to continue surveying their presence in surface waters and in wastewater treatment
Surface water plant (WWTP) effluents.
Ecotoxicity Furthermore, due to the rise of anticancer home treatments, most of the prescribed molecules are now available
Occurrence in town pharmacies. Consequently, hospital effluents are no longer the main expected entry route of anticancer
drugs into the aquatic environment.
Concerning ecotoxicological risks, current knowledge remains insufficient to support a definitive conclusion.
Risk posed by cytotoxic molecules is still not well documented and it is not possible to conclude on their
long-term effects on non-target organisms. To date, ecotoxicological effects have been assessed using standardized
or in vitro assays. Such tests however may not be suitable for anticancer drugs, and further work should focus on
full-life cycle or even multigenerational tests.
Environmental significance (i.e. occurrence and effects) of cytostatics (protein kinases inhibitors and monoclonal
antibodies), if any, is not documented. Protein kinases inhibitors, in particular, deserve further investigation due
to their universal mode of action.
Finally, concerning endocrine therapy drugs, molecules such as antiestrogen Tamoxifen and its active metabolites,
could be of concern.
Overall, to accurately assess the ecotoxicological risk of anticancer drugs, we discuss the need to break away from
tests on isolated molecules and to test effects of mixtures at the low ng.l− 1 range.
© 2011 Elsevier Ltd. All rights reserved.

1. Introduction
Since their first detection in the aquatic environment in the 1970s
(Hignite and Aznaroff, 1977), growing attention has been paid to phar-
maceuticals. As these compounds reach the environment in non-
⁎ Corresponding author. Tel.: + 33 4 72 20 87 87, + 33 4 72 20 89 05; fax: + 33 4 78
47 78 75.
E-mail addresses: jean-philippe.besse@cemagref.fr (J.-P. Besse),
latour@lyon.fnclcc.fr (J.-F. Latour), jeanne.garric@cemagref.fr (J. Garric).
1
Tel.: + 33 4 78 78 28 28.
2
Tel.: + 33 4 72 20 87 87; fax: + 33 4 78 47 78 75.
negligible amounts and because they are likely to exert biological and/
or adverse effects on aquatic organisms, they can be considered as envi-
ronmental contaminants. Subsequently, there is a need to target phar-
maceuticals with environmental significance, quantify them, and
assess their ecotoxicological risk (Fent et al., 2006; Halling-Sørensen
et al., 1998; Kolpin et al., 2002; Ternes, 1998). Among pharmaceuticals,
drugs used for cancer treatment (referred to as anticancer or antineo-
plastic drugs) are suspected to represent a specific risk for aquatic
non-target species (Kümmerer, 2001).
Currently, two main groups of drugs are used: cytotoxic drugs and
endocrine therapy drugs. Due to their intended function, the first ones
could exert cytotoxic, genotoxic, mutagenic, carcinogenic or teratogenic
effects on aquatic species. Moreover, as they often act with structure

0160-4120/$ – see front matter © 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.envint.2011.10.002
74 J.-P. Besse et al. / Environment International 39 (2012) 73–86

Table 1
A classification of anticancer drugs (cytotoxic, cytostatic and hormonally active drugs).

ATC class Class Mechanism of action Detailed mechanism

Antineoplastic Cytotoxics which interact
agents (ATC L01) directly with DNA
Cytotoxics which interact
indirectly with DNA
Cytostatics
Endocrine therapy Hormones and related
(ATC L02) Anti-hormones and related
Alkylating agents
Platinum complexes
Intercalating agents
Antimetabolites
Cytotoxic antibiotics
Mitotic spindle inhibitors
Topoisomerase inhibitors
Protein kinase inhibitors
Monoclonal antibodies
GnRH and LhRH analogs
Antiestrogens
Antiaromatase
Antiandrogens
Attach one or more nucleophile groups to the DNA
and inhibit or alter its transcription
Inhibit DNA replication by binding to it
Act by breaking single-stranded DNA
Structural analogs of purine, pyrimidine or folic acid,
act by blocking enzyme activity and disrupting DNA synthesis
Intercalate between DNA base pairs and thus disturb the synthesis
and/or function of nucleic acids
Halt chromosome segregation by inhibiting mitotic spindle formation
Induce or stabilize DNA damage by blocking religation of double stranded DNA breaks
Interact with protein kinases, involved in the regulation of many
biological processes (cell growth, migration,survival..),
most of the compounds used are Tyrosine kinase inhibitors
Block tumoral cells extracellular receptors
Interfere with gonadotropin release
Direct estrogen receptor antagonism
Inhibit the estrogen biosynthesis
Inhibit the binding of testosterone to its receptor

and function of DNA, some authors consider that all eukaryote organisms
might be susceptible to their toxicity (Johnson et al., 2008).
Drugs used in endocrine therapy, due to their specific hormonal or
anti-hormonal properties can be considered as endocrine disruptors
and therefore require consideration.
Overall, considering their distinctive mechanisms of action, anti-
cancer drugs should be screened and assessed for environmental
risk, according to the EMEA guidelines (EMEA, 2006). However, as
the EMEA guidelines only require environmental assessment for
newly authorized pharmaceuticals, most of the anticancer drugs are
not submitted to such a procedure.
Some authors have questioned the environmental occurrence and
risk of cytotoxic (Henschel et al., 1997; Johnson et al., 2008; Kümmerer
and Al-Ahmad, 1997; Straub, 2010; Zounkova et al., 2010), and endo-
crine therapy drugs (Runnalls et al., 2010). Others have modeled their
release in hospital effluents (Mullot, 2009; Mullot et al., 2010). Never-
theless, it appears that only a few drugs have been assessed and that
no focused effort has been made to ascertain which molecules are likely
to pose a risk to the aquatic environment.
This paper therefore aims at achieving three goals:
• providing an overview of the occurrence of anticancer drugs in
the aquatic environment by calculating predicted environmental
concentrations (PECs), based on French consumption data;
• conducting a preliminary risk characterization for these molecules
and give hints for further ecotoxicological testing;
• investigating consumption trends.
In this study, only the risk for surface waters is discussed. Risk
related to drinking waters has been discussed by other authors
(Johnson et al., 2008; Rowney et al., 2009); risk associated with
sludge and sediment is not well documented, however some existing
data suggest that hydrophobic anticancers and platinum derivatives
may partition on particles (Lenz et al., 2005; Thomas et al., 2007);
finally, risk related to groundwater, if any, has not yet been documented.
structure and mechanism of action: cytotoxics, strictly speaking, and
cytostatics.
Cytotoxic drugs are compounds that interact with DNA. They
cause metabolic and morphological alterations of the cell, leading to
its death, through different mechanisms of action (Table 1); they
are reported to exert genotoxic, mutagenic and carcinogenic effects.
Cytostatic agents display mechanisms of action somewhat different,
leading to a decrease in tumoral cell proliferation.
Cytostatics do not interact with DNA and can achieve their effects
through various mechanisms, such as blocking cell growth factors, or
indirectly by recruiting cytotoxic cells, such as macrophages and
monocytes, which will contribute to the tumoral cell death. Cytostatics
include monoclonal antibodies (ATC class L01XC) and anti-protein
kinases (ATC class L01XE); most of the protein kinase inhibitors
used are tyrosine kinase inhibitors (Table 1).
Aside from these two classes, other molecules are used for treatment
of hormone-dependant cancers. They are referred to as endocrine
therapy drugs, and are classed into two groups: hormone analogs and
related agents (L02A), and hormone antagonists and related agents
(L02B).
Hormone analogs used in cancer treatment are molecules that
interfere with gonadotropin release. These molecules are either
GnRH analogs or LhRH analogs. 3 Hormone antagonists include several
non-steroidal antiestrogens and antiandrogens: antiestrogens act either
by inhibiting the estrogen biosynthesis (antiaromatase compounds) or
by direct estrogen receptor antagonism (Table 1). Antiandrogens are
compounds with a non-steroidal structure that inhibit the binding of
testosterone to its receptor (Table 1).
For convenience, the term anticancer drug will be used throughout
this document when accounting for the all types of molecules
described above (cytotoxics, cytostatics and endocrine therapy).
3. Materials and methods

2. Classification of anticancer drugs, cytotoxic, cytostatic and
endocrine therapy drugs
In studies conducted on occurrence and effects of anticancer drugs
in the environment, the terms cytotoxic or cytostatic are both used
without discrimination to account for anticancer drugs in general.
However, cytotoxic and cytostatic agents can differ greatly in their
mechanism of action.
Cytotoxic drugs belong to the L01 class of the Anatomical Therapeutic
Chemical (ATC) classification (www.whocc.no/atcddd). Cytotoxic agents
can be classified in two main groups, on the basis of their chemical
To gather and investigate pharmacological, metabolism and
ecotoxicological data for anticancer compounds, scientific literature
was reviewed as along with the following databases and books: the
Banque Claude Bernard (BCB, 2011), the Micromedex Drugdex®
databank (Micromedex Drugdex©, 2011), the IARC database
(www.iarc.fr), and the BC cancer agency database (www.bccancer.
3
Some progestagens, namely progesterone and medroxyprogesterone, can be used
in the treatment of some cancers. For these molecules, expected concentrations and ef-
fects on the aquatic environment have been discussed elsewhere (Besse and Garric,
2009).
 J.-P. Besse et al. / Environment International 39 (2012) 73–86
bc.ca), and the Drug bank database (www.drugbank.ca). Gathered
data on metabolism are provided in Appendix A.
Preliminary exposure assessment was implemented by calculating
PECs, as described by Besse et al. (2008), using the following Eq. (1):
amount  Fexcreta  Fstp
PEC ¼ ð1Þ
WWinhab  hab  Dilution  365
PECs are expressed in mg.l − 1 using the following parameters:
consumption is the quantity (mg.year − 1) of an active molecule
consumed by the population over 1 year in a defined zone (generally
a country); hab is the number of inhabitants and 100 the correction
factor for the percentages; 365 is the number of days per year
(day.year − 1); WWinhab is the volume of wastewater per person
per day (default value = 200 l.inhabitants − 1.day − 1); dilution is the
dilution factor from wastewater treatment plant (WWTP) effluents
to surface waters (default value set at 10); Fexcreta is the excretion
fraction of the active molecule; Fstp is the fraction of emission of
the drug from WWTPs directed to surface water, which can be
defined as (1-WWTP removal fraction). In most cases, WWTP removal
rates were not available and therefore we assumed an Fstp value of 1,
which corresponds to a worst-case scenario (i.e. no removal in
WWTPs).
Calculated PECs and investigated consumption trends are based
on national and local data sets. At the national scale, French actual
amounts, kindly provided by the French Health Products Safety Agency
(Agence Française de Sécurité Sanitaire des Produits de Santé, AFSSAPS,
Paris) were used: two data sets for cytotoxic (ATC L01) compounds for
years 2004 and 2008 (respectively AFSSAPS, 2006; and AFSSAPS,
2009a); and one data set for endocrine therapy drugs (ATC L02) for
year 2008 (AFSSAPS, 2009b).
At the local scale, consumption data for years 2004 and 2008,
originating from a regional specialized cancer care centre, the Centre
Léon Bérard (CLB) located in Lyon, France, were used.
Gathered data are given in Appendices B and C, respectively for
AFSSAPS and CLB data.
4. Results
4.1. Exposure assessment
Based on Eq. (1), PEC values were determined for a wide range of anticancer drugs.
The calculated values are presented in Table 2 for years 2004 and 2008. PECs proposed
here provide a rough insight of the national contamination, not accounting for local
specificities.
Due to scarcity of data, WWTP removal rates were not taken into account in the
following calculations. Two PEC values were determined for each molecule: a conservative
PEC, assuming no human metabolism, and a second one, refined by excretion rates, when
available.
Conservative PECs are to be considered as a representation of what currently enters
the aquatic environment rather than what is in reality occurring in surface water. On
the basis of such a worst-case scenario (i.e. 100% excretion and 0% removal in
WWTPs), calculated values appear to be very low, with only 6 molecules displaying a
conservative PEC higher than 10 ng.l− 1 (the EMEA threshold value for the environmental
risk assessment of pharmaceuticals; EMEA, 2006) and 22 molecules with a PEC higher
than 1 ng.l− 1. Overall, nearly half of the anticancers used display very low conservative
PECs, lower than 0.1 ng.l− 1 (Table 2). Compared to most of the other pharmaceutical
classes, conservative PECs for anticancer drugs are much lower. They are up to 3 orders of
magnitude lower than anti-inflammatory drugs, antihypertensives and certain antibiotics.
They are however in the lower range of serotoninergic antidepressants (Besse et al.,
2008), and in the same range than synthetic estrogens and progestagens (Besse and Garric,
2009).
For cytotoxics, only three molecules have a conservative PEC greater than 10 ng.l− 1:
Hydroxycarbamide, Capecitabine and Fluorouracil. For cytostatic compounds, Imatinib is
the only molecule displaying a conservative value higher than 10 ng.l− 1, while seven
molecules have a PEC higher than 1 ng.l− 1 (Table 2). Finally, for endocrine therapy
drugs, calculated values are also under the ng.l− 1 range, except for a few molecules:
the antiandrogens Bicalutamide, Flutamide (PEC higher than 10 ng.l− 1); the antiandrogen
Nilutamide, the antiestrogen Tamoxifen and the antiaromatase Exemestane (PECs higher
than 1 ng.l− 1).
Metabolism data were retrieved for approximately one-third of the anticancer drugs
investigated here; interestingly, such data are mainly available for the most consumed
(and therefore the most studied) drugs. Excretion rates vary from less than 5% up to
75
100%; 12 molecules have excretion rate values equal to or lower than 10%, while 14
molecules have values greater than or equal to 50%. PECs refined by metabolism are
given in Table 2.
A recent study (Rowney et al., 2009) calculated PEC values for 10 cytotoxic molecules
in WWTP influent and effluent in England. If converted to values for WWTP effluent (i.e.
without the dilution factor of 10), our refined PEC are close to those calculated in this
study, except for Fluorouracil, for which our PEC is one order of magnitude higher. Such
a discrepancy is notably linked to the fact that this molecule is highly removed in
WWTPs (Mahnik et al., 2004; Rowney et al., 2009; Straub, 2010), which was not taken
into account in our calculation.
Only a few cytotoxics have been searched and/or detected in the aquatic environment,
and concentrations are generally very low, in the ng.l− 1 range or below detection limits
(Table 3). As WWTP removal rates were not taken into account, calculated PECs for
cytotoxics appear consistent with measured concentrations in wastewaters, but are
generally higher than measured levels in surface waters.
For cytostatics, to our knowledge, no occurrence data are available for these
molecules, as environmental research has not yet focused on this class of molecules.
Among the investigated endocrine therapy drugs, only a few have been searched in
the environment (Table 3). Tamoxifen is reported with maximum concentrations in
WWTP effluent of 42 ng.l− l (Ashton et al., 2004), and of 102 ng.l− l and of 25 ng.l− l in
WWTP effluent and in surface water respectively (Coetsier et al., 2009); nevertheless,
in the bulk of measurements, concentrations are below the detection limit. For
other compounds, only data from Liu et al. (2010) are available, with Letrozole and
Anastrozole measured with a high frequency, and with maximum respective concentrations
of 2.38 and 3.70 ng.l− l in hospital effluent, and of 0.3 and 0.4 ng.l− l in WWTP
effluent (Table 3). Only conservative PECs could be determined for these molecules,
calculated values for Letrozole and Anastrozole are one order of magnitude higher
than the reported ones whereas calculated value for Tamoxifen is in the range of
field measurements (Table 2).
4.2. Consumption trends for cytotoxic and cytostatic compounds
Data from the French Health Products Safety Agency for cytotoxic and cytostatic
drugs (AFSSAPS, 2006 and AFSSAPS, 2009a) show that national consumption amounts
have increased between 2004 and 2008 (Fig. 1a and b); total amounts reached 13 t in
2004 and have increased to 17.5 t in 2008 (Appendix B).
Results also show that antimetabolites, mainly represented by Capecitabine and
Fluorouracil (90% of the total amounts of antimetabolites in 2008), and Hydroxycarbamide
(Hydroxyurea) are the main classes used in France (Fig. 1a and b). Hydroxycarbamide
(ATC class L01XX05) is by far the most consumed anticancer drug in France. Its consump-
tion represents 44% and 39% of the total amounts for 2004 and 2008 respectively, and it is
almost completely available in town pharmacies. Due to this specific consumption pattern,
we chose to consider this molecule separately from others. Antimetabolites and alkylating
agents are the main classes used in hospitals (Fig. 1a and b; Appendix B).
If consumed amounts have increased for almost all anticancer chemical classes,
except for plant alkaloids (Fig. 1b; Appendix B), this is not the case for all individual
drugs: for example, Mitotane consumption has doubled over the past 4 years, whereas a
nearby 40% decline in consumption amounts can be observed for Methotrexate.
Consumption trends also indicate that some molecules have completely disappeared,
replaced by new ones, presumably more efficient or with less adverse effects. To this
extent, it can be observed that the use of cytostatic molecules (protein kinase inhibitors
and monoclonal antibodies) has increased from 2004 to 2008 and that consumption
amounts have risen for all the compounds belonging to this family (from 4% to 8% of
the total amounts). Among cytostatics, Protein kinase inhibitors are almost completely
delivered in town pharmacies (Fig. 1b).
The most interesting point however, is the distinct shift that can be observed in
consumption patterns between 2004 and 2008 (Fig. 1a and b; Appendix B). In 2004,
82% of anticancer drugs were delivered in hospitals (regardless of Hydroxycarbamide,
see Appendix B for details), while in 2008 this proportion was only about 35%. This
shift is observed for almost all classes of anticancer drugs, and notably for antimetab-
olites (Fig. 1a and b). This is of importance for the assessment of entry routes of anti-
cancers in the aquatic environment (see Section 5.2).
Local data from the Centre Léon Bérard (Lyon, France) also show that consump-
tion amounts have increased between 2004 and 2008 (Fig. 2a and b). CLB total
amounts consumed were of 22 kg in 2004 and have increased to 26 kg in 2008 (Ap-
pendix C).
Not all the molecules used at the national scale are used at the CLB (Appendix C).
Nevertheless the CLB local consumption trends are similar to the national ones, and the
main chemical classes used are antimetabolites and alkylating agents, which confirms
observations made at the national scale for hospitals (Fig. 2a; 1a and 1b). As Hydroxy-
carbamide is almost completely delivered in town pharmacies, it is logically used in
very low amounts at the CLB (Appendix C). Contrary to national trends, the use of
plant alkaloids has increased between 2004 and 2008 at the CLB. Overall, and despite
differences observed for a limited number of molecules (probably linked to local patterns
in chemotherapy protocols), a good correlation is observed between national and the CLB
local data for cytotoxic molecules (Appendix C).
Concerning cytostatic molecules, the CLB data show that the use of monoclonal
antibodies has considerably risen between 2004 and 2008 as consumed amounts
have increased by a factor of 3 (Fig. 2a; Appendix C). Protein kinase inhibitors are
not used at all at the CLB, which confirms observations made at the national scale.
76 J.-P. Besse et al. / Environment International 39 (2012) 73–86

Table 2
Predicted environmental concentrations (PECs) for anticancer drugs, based on their consumption in France for years 2004 and 2008 (data from AFSSAPS, 2006, 2009a, 2009b).
Concentrations expressed in ng.l− 1. Molecules are sorted by decreasing consumption amounts (for year 2008). WWTP removal rates were not taken into account in the calculation of
the following values.

Molecule Type of Chemical class Total amounts Total amounts Conservative PEC Conservative PEC F excreta Refined PEC Refined PEC
molecule 2004 (kg) 2008 (kg) 2004 (ng l− 1) 2008 (ng l− 1) 2004 (ng.l− 1) 2008 (ng.l− 1)

Hydroxycarbamide Cytotoxic Other 5756.67 6838.63 131.43 156.13 0.50 65.72 78.07
Capecitabine Cytotoxic Antimetabolite 2620.99 5134.94 59.84 117.24 0.03 1.80 3.52
Fluorouracil Cytotoxic Antimetabolite 1690.24 1733.20 38.59 39.57 0.20 7.72 7.91
Imatinib Cytostatic Tk inhibitor 583.68 873.90 13.33 19.95 0.25 3.33 4.99
Bicalutamide Endocrine Antiandrogen n.a. 863.00 n.a. 19.70 0.55 n.a. 10.84
Flutamide Endocrine Antiandrogen n.a. 521.00 n.a. 11.90 b 0.10 n.a. b 1.19
Gemcitabin Cytotoxic Antimetabolite 339.21 379.28 7.74 8.66 0.10 0.77 0.87
Tamoxifen Endocrine Antiestrogen n.a. 377.00 n.a. 8.61 n.a. n.a. n.a.
Cyclophosphamide Cytotoxic Alkylating agent 281.84 305.73 6.43 6.98 > 0.25 > 1.61 >1.74
Estramustine Cytotoxic Other 388.38 287.62 8.87 6.57 0.05 0.44 0.33
Mitotane Cytotoxic Other 95.90 233.75 2.19 5.34 0.60 1.31 3.20
Exemestane Endocrine Antiaromatase n.a. 182.00 n.a. 4.16 n.a. n.a. n.a.
Nilutamide Endocrine Antiandrogen n.a. 169.00 n.a. 3.86 b 0.03 n.a. b0.12
Erlotinib Cytostatic Tk inhibitor 0.00 148.85 0.00 3.40 b 0.02 0.00 b0.07
Cytarabine Cytotoxic Antimetabolite 117.41 133.59 2.68 3.05 b 0.10 b 0.27 b0.31
Lapatinib Cytostatic Tk inhibitor 0.00 116.20 0.00 2.65 0.70 0.00 1.86
Ifosfamide Cytotoxic Alkylating agent 121.38 103.04 2.77 2.35 0.50 1.39 1.18
Mercaptopurine Cytotoxic Antimetabolite 102.04 94.84 2.33 2.17 0.07 0.16 0.15
Bevacizumab Cytostatic Monoclonal ab 0.00 87.12 0.00 1.99 n.a. n.a. n.a.
Carboplatin Cytotoxic Platinum derivative 64.38 83.59 1.47 1.91 1.00 1.47 1.91
Methotrexate Cytotoxic Antimetabolite 117.63 74.73 2.69 1.71 0.90 2.42 1.54
Rituximab Cytostatic Monoclonal ab 32.52 72.08 0.74 1.65 n.a. n.a. n.a.
Pipobroman Cytotoxic Alkylating agent 63.06 66.93 1.44 1.53 n.a. n.a. n.a.
Nilotinib Cytostatic Tk inhibitor 0.00 58.71 0.00 1.34 0.60 0.00 0.80
Trastuzumab Cytostatic Monoclonal ab 15.05 56.01 0.34 1.28 n.a. n.a. n.a.
Cetuximab Cytostatic Monoclonal ab 7.38 55.03 0.17 1.26 n.a. n.a. n.a.
Temozolomide Cytotoxic Alkylating agent 29.23 53.54 0.67 1.22 0.10 0.07 0.12
Irinotecan Cytotoxic Other 33.89 46.58 0.77 1.06 > 0.50 > 0.38 >0.53
Etoposide Cytotoxic Plant alcaloid 332.84 41.11 7.60 0.94 0.93 7.07 0.87
Tegafur Cytotoxic Antimetabolite 86.51 37.31 1.98 0.85 n.a. n.a. n.a.
Paclitaxel Cytotoxic Plant alcaloid 27.29 38.75 0.62 0.88 0.18 0.11 0.16
Pemetrexed Cytotoxic Antimetabolite 0.92 37.31 0.02 0.85 0.90 0.02 0.77
Procarbazine Cytotoxic Methylhydrazine 16.48 34.55 0.38 0.79 b 0.20 b 0.08 b0.16
Letrozole Endocrine Antiaromatase n.a. 34.30 n.a. 0.78 n.a. n.a. n.a.
Oxaliplatin Cytotoxic Platinum derivative 20.34 33.47 0.46 0.76 n.a. n.a. n.a.
Anastrozole Endocrine Antiaromatase n.a. 31.70 n.a. 0.72 n.a. n.a. n.a.
Dacarbazine Cytotoxic Alkylating agent 18.65 29.45 0.43 0.67 0.50 0.22 0.34
Docetaxel Cytotoxic Plant alcaloid 16.41 27.47 0.37 0.63 b 0.08 b 0.03 b0.05
Bexarotene Cytotoxic Other 8.24 23.62 0.19 0.54 n.a. n.a. n.a.
Cisplatin Cytotoxic Platinum derivative 17.31 22.74 0.40 0.52 n.a. n.a. n.a.
Sunitinib Cytostatic Pk inhibitor 0.00 20.10 0.00 0.46 n.a. n.a. n.a.
Epirubicin Cytotoxic Cytotoxic ATB 18.80 17.59 0.43 0.40 n.a. n.a. n.a.
Doxorubicin Cytotoxic Cytotoxic ATB 7.94 16.82 0.18 0.38 0.50 0.09 0.19
Vinorelbine Cytotoxic Plant alcaloid 6.29 12.97 0.14 0.30 0.11 0.02 0.03
Streptozocin Cytotoxic Cytotoxic ATB 7.43 8.38 0.17 0.19 0.10 0.02 0.02
Chlorambucil Cytotoxic Alkylating agent 6.65 8.18 0.15 0.19 n.a. n.a. n.a.
Fulvestrant Endocrine Antiestrogen n.a. 6.70 n.a. 0.15 n.a. n.a. n.a.
Fludarabine Cytotoxic Antimetabolite 3.13 5.51 0.07 0.13 n.a. n.a. n.a.
Melphalan Cytotoxic Alkylating agent 3.80 4.86 0.09 0.11 n.a. n.a. n.a.
Tretinoin Cytotoxic other 1.84 3.28 0.04 0.07 n.a. n.a. n.a.
Leuprorelin Endocrine Lh-RH analog n.a. 3.12 n.a. 0.07 n.a. n.a. n.a.
Lomustine Cytotoxic Cytotoxic ATB 0.37 3.09 0.01 0.07 n.a. n.a. n.a.
Altretamin Cytotoxic Other 3.21 3.06 0.07 0.07 n.a. n.a. n.a.
Mitomycin C Cytotoxic Cytotoxic ATB 2.04 3.01 0.05 0.07 > 0.10 > 5.00 E-3 >7.00 E-3
Aminolevulinate Cytotoxic Photodynamic th. 0.00 2.39 0.00 0.05 n.a. n.a. n.a.
Thioguanine Cytotoxic Antimetabolite 1.93 2.23 0.04 0.05 n.a. n.a. n.a.
Triptorelin Endocrine Lh-RH analog n.a. 2.00 n.a. 0.05 n.a. n.a. n.a.
Carmustine Cytotoxic Cytotoxic ATB 1.41 1.81 0.03 0.04 n.a. n.a. n.a.
Mitoguazone Cytotoxic Other 1.03 1.68 0.02 0.04 n.a. n.a. n.a.
Fotemustine Cytotoxic Cytotoxic ATB 1.46 1.31 0.03 0.03 n.a. n.a. n.a.
Panitumumab Cytostatic Monoclonal ab 0.00 1.28 0.00 0.03 n.a. n.a. n.a.
Anagrelide Cytotoxic Other 0.00 1.23 0.00 0.03 n.a. n.a. n.a.
Temsirolimus Cytostatic Pk inhibitor 0.00 1.19 0.00 0.03 n.a. n.a. n.a.
Goserelin Endocrine GnRH analog n.a. 1.14 n.a. 0.03 n.a. n.a. n.a.
Toremifen Endocrine Antiestrogen n.a. 1.00 n.a. 0.02 n.a. n.a. n.a.
Bleomycine Cytotoxic Cytotoxic ATB 0.72 0.88 0.02 0.02 n.a. n.a. n.a.
Daunorubicin Cytotoxic Cytotoxic ATB 0.77 0.76 0.02 0.02 n.a. n.a. n.a.
Alemtuzumab Cytostatic Monoclonal ab 0.00 0.72 0.00 0.02 n.a. n.a. n.a.
Vinblastine Cytotoxic Plant alcaloid 0.32 0.71 0.01 0.02 n.a. n.a. n.a.
Amsacrine Cytotoxic Other 0.16 0.39 4.00 E-3 9.00 E-3 0.20 0.80 E-3 1.80 E-3
Thiotepa Cytotoxic Alkylating agent 0.21 0.35 5.00 E-3 8.00 E-3 n.a. n.a. n.a.
Mitoxantrone Cytotoxic Cytotoxic ATB 0.29 0.27 7.00 E-3 6.00 E-3 n.a. n.a. n.a.
 J.-P. Besse et al. / Environment International 39 (2012) 73–86 77

Table 2 (continued)
Molecule Type of Chemical class Total amounts Total amounts Conservative PEC Conservative PEC F excreta Refined PEC Refined PEC
molecule 2004 (kg) 2008 (kg) 2004 (ng l− 1) 2008 (ng l− 1) 2004 (ng.l− 1) 2008 (ng.l− 1)

Miltefosine Cytotoxic Other 0.00 0.22 0.00 5.00 E-3 n.a. n.a. n.a.
Bortezomib Cytotoxic Other 0.00 0.22 0.00 5.00 E-3 n.a. n.a. n.a.
Topotecan Cytotoxic Other 0.10 0.21 2.00 E-3 5.00 E-3 n.a. n.a. n.a.
Idarubicine Cytotoxic Cytotoxic ATB 0.13 0.17 3.00 E-3 4.00 E-3 n.a. n.a. n.a.
Vincristine Cytotoxic Plant alcaloid 0.14 0.15 3.00 E-3 3.00 E-3 n.a. n.a. n.a.
Clofarabine Cytotoxic Antimetabolite 0.00 0.14 0.00 3.00 E-3 n.a. n.a. n.a.
Pirarubicine Cytotoxic Cytotoxic ATB 0.03 0.13 1.00 E-3 3.00 E-3 n.a. n.a. n.a.
Vindesine Cytotoxic Plant alcaloid 0.05 0.06 1.00 E-3 1.00 E-3 n.a. n.a. n.a.
Arsenic trioxyde Cytotoxic Other 0.02 0.05 1.00 E-3 1.00 E-3 n.a. n.a. n.a.
Buserelin Endocrine GnRH analog n.a. 0.05 n.a. 1.00 E-3 n.a. n.a. n.a.
Porfimer sodium Cytotoxic Photodynamic th. 0.01 0.04 0.00 1.00 E-3 n.a. n.a. n.a.
Cladribine Cytotoxic Antimetabolite 0.02 0.03 0.00 1.00 E-3 n.a. n.a. n.a.
Busulfan Cytotoxic Alkylating agent 0.15 0.03 1.00 E-3 1.00 E-3 n.a. n.a. n.a.
Raltitrexed Cytotoxic Antimetabolite 0.03 0.03 1.00 E-3 1.00 E-3 0.50 0.50 E-3 0.50 E-3
Pentostatine Cytotoxic Other 0.00 0.02 0.00 1.00 E-3 n.a. n.a. n.a.
Alitretinoin Cytotoxic Other 0.00 18.00 E-3 0.00 4.00 E-4 n.a. n.a. n.a.
Trabectedine Cytotoxic Plant alcaloid 0.00 13.00 E-3 0.00 3.00 E-4 n.a. n.a. n.a.
Chlormetine Cytotoxic Alkylating agent 0.56 8.00 E-3 1.00 E-3 2.00 E-4 n.a. n.a. n.a.
Ibritumomab tiuxetan Cytotoxic Photodynamic th. 78.00 3.00 E-3 1.78 1.10-4 n.a. n.a. n.a.

Tk inhibitor: Tyrosine kinase inhibitor.

Pk inhibitor: Protein kinase inhibitor.
Monoclonal ab: Monoclonal antibody.
Cytotoxic ATB: Cytotoxic antibiotic.
Photodynamic th.: Photodynamic therapy.
n.a.: not available or incomplete data.
Fexcreta: fraction excreted as unchanged drug.

Table 3
Concentrations of anticancer drugs in different environmental samples (expressed in ng.l− 1). WWTP : wastewater treatment plant; nd: not detected; m: mean; md: median; max:
maximum.

Molecule Sample Concentration (ng.l− 1) References

Cyclophosphamide Surface water nd–64.8 Moldovan (2006)

WWTP effluent 0.6 (md) Zuccato et al. (2005)
Hospital effluent 20–4500 Steger-Hartmann et al. (1996)
WWTP influent nd–143 Steger-Hartmann et al. (1996)
WWTP effluent nd–17 Steger-Hartmann et al. (1996)
Surface water 0.05–0.17 Buerge et al. (2006)
Hospital effluent 30–900 Catastini et al. (2008)
WWTP effluent 300 Catastini et al. (2008)
Ifosfamide Hospital effluent nd–1914 (md = 109) Kümmerer et al. (1997)
WWTP influent nd–40 Kümmerer et al. (1997)
WWTP influent nd (md)–2900 (max) Ternes (1998)
Surface water nd (md)–20 (max) Ternes (1998)
Surface water b0.05–0.14 Buerge et al. (2006)
Hospital effluent 30–900 Catastini et al. (2008)
WWTP effluent 100 Catastini et al. (2008)
Hospital effluent b2–10,647 (max) Yin et al. (2010)
Daunorubicin Hospital effluent 260–1350 Mahnik et al. (2007)
Etoposide Hospital effluent 110–600 Catastini et al. (2008)
Hospital effluent b5–380 (max) Yin et al. (2010)
Methotrexate Hospital effluent nd–200 Catastini et al. (2008)
WWTP effluent 30 Catastini et al. (2008)
Hospital effluent b2–4689 (max) Yin et al. (2010)
Bleomycin WWTP influent 15.8 (m) Aherne et al. (1990)
Surface water 8.5 (m) Aherne et al. (1990)
Hospital effluent b8600–124,000 Mahnik et al. (2004)
Hospital effluent nd–30 Catastini et al. (2008)
Tamoxifen Surface water 27–212 Roberts and Thomas (2006)
WWTP effluent nd–42 (max) Ashton et al. (2004)
WWTP effluent 102 (max) Coetsier et al. (2009)
Surface water 25 (max) Coetsier et al. (2009)
Hospital effluent 0.2–8.2 Liu et al. (2010)
WWTP influent 0.28 Liu et al. (2010)
Letrozole Hospital effluent 0.20–2.38 Liu et al. (2010)
WWTP influent 0.28–0.8 Liu et al. (2010)
WWP effluent 0.27–0.6 Liu et al. (2010)
Anastrozole Hospital effluent 0.3–3.7 Liu et al. (2010)
WWTP influent 0.12–0.32 Liu et al. (2010)
WWP effluent 0.3 Liu et al. (2010)
78 J.-P. Besse et al. / Environment International 39 (2012) 73–86

a a
8000 9000

7000 8000
7000
6000
6000
5000
5000
kg 4000 g
4000
3000 3000
2000 2000
1000 1000
0 0
Antimetabolites Hydroxycarbamide Alkylating agents Antimetabolites
Pharmacies 2004 Pharmacies 2008 Hospital 2004 Hospital 2008
Complete hosp 2004 Complete hosp 2008 Outpatients 2004 Outpatients 2008

b b
1200
1400
1000
1200
800
1000
kg 600
800
400 g
600
200
400
0
Alkylating Alcaloids and Cytotoxic Platinum Monoclonal Protein Other 200
agents related antibiotics derivatives antibodies kinase
inhibitors 0
Pharmacies 2004 Pharmacies 2008 Hospital 2004 Hospital 2008 Alcaloids and Cytotoxic Platinum Monoclonal Other
related antibiotics derivatives antibodies
Fig. 1. a. Consumption trends in French town pharmacies and hospitals for years 2004 Complete hosp. 2004 Complete hosp 2008 Outpatients 2004 Outpatients 2008
and 2008, for the two most consumed anticancer chemical classes (data from AFSSAPS,
2006, 2009a). Amounts are given in kg. Due to its high consumption amounts, Fig. 2. a. Consumption trends for anticancer drugs at the Centre Léon Bérard (Lyon,
Hydroxycarbamide was treated separately from other molecules (see Section 4.2). France); sorted by “Complete hospitalization” and “Outpatients” for years 2004 and
b. Consumption trends in French town pharmacies and hospitals for years 2004 and 2008 for the two most consumed anticancer chemical classes (CLB data; Appendix C).
2008, for the other anticancer chemical classes (data from AFSSAPS, 2006, 2009a). Amounts are given in grams. Complete hosp: Complete hospitalization. b. Consumption
Amounts are given in kg. Due to its high consumption amounts, Hydroxycarbamide trends for anticancer drugs at the Centre Léon Bérard (Lyon, France); sorted by
was treated separately from other molecules (see Section 4.2). “Complete hospitalization” and “Outpatients” for years 2004 and 2008 for the other
anticancer chemical classes (CLB data; Appendix C). Amounts are given in grams.
Complete hosp: Complete hospitalization.

Results are displayed in Table 4. Preferential molecules are:

5. Discussion
5.1. Targeting anticancer drugs to be searched in surface waters
5.1.1. Parent molecules
For anticancer drugs, scientific knowledge is limited, both on
ecotoxicological and occurrence data. Consequently, and pragmatically,
in order to gain knowledge on these molecules, it seems sound to first
focus on compounds which are the most likely to be present in effluents
or surface waters: indeed, searching for anticancer contaminants at the
ng.l− 1 range means facing analytical challenges. Therefore, as a first
step, we consider that it is better to focus primarily on molecules for
which there is a higher probability of detection in the environment.
From this perspective, we built a “preferential” (rather than priority)
list, based on exposure only, which was implemented as follows:
Considering the range of measured concentrations in the environ-
ment, the number of molecules with a conservative PEC lower than
1 ng.l− 1, and the analytical feasibility, we chose arbitrarily to set a
threshold value at 1 ng.l− 1 in order to rank anticancer drugs. Molecules
with a conservative PEC (i.e. without taking into account metabolism
and WWTP removal rates) lower than this value were a priori excluded
from the list of preferential molecules.
Among compounds with a PEC higher than 1 ng.l− 1, metabolism,
WWTP removal rates and biodegradation data were reviewed exhaus-
tively, and refined PECs were calculated when possible (using Eq. (1)).
Molecules with a refined PEC higher than 1 ng.l− 1 were considered as
preferential compounds. For compounds with a conservative PEC
higher than 1 ng.l− 1 and for which neither reliable excretion data nor
reliable biodegradation data were available, we considered that it was
not possible to conclude.
Hydroxycarbamide, Capecitabine, Cyclophosphamide, Ifosfamide,
Mitotane, Imatinib, Lapatinib, Tamoxifen and Flutamide. For each
molecule, its inclusion to the list is discussed along with subsequent
recommendations.
Even though irrelevant to the standpoint of the risk, such a list is
useful to i) help to define environmental tracers of anticancers
contamination, ii) highlight molecules on which further occurrence
and ecotoxicological studies should be conducted and iii) determine, for
the implementation of ecotoxicological assays, exposure concentration
ranges consistent with the environmental reality.
5.1.2. Metabolites
As in previous works (Besse and Garric, 2008; Besse et al., 2008),
special attention was paid to metabolites. Metabolism of anticancer
compounds is complex and leads to the formation of multiple
pharmacologically active and inactive metabolites. However, if antican-
cer drug metabolism is well documented (Appendix A), quantitative
available data (i.e. excretion rates) are limited and only allow calculating
PECs for a very few metabolites.
A preferential list was established for metabolites following the
same rules than for parent compounds. Metabolites considered here
were those reported to be pharmacologically active and/or those
with an excretion rate higher than 10% (Besse et al., 2008; EMEA,
2006). Results and discussions are presented in Table 5. For metabolites,
degradation data and environmental behavior are more than limited, so
it remains uncertain to which extent they can occur in the aquatic
environment.
Here we briefly discuss a few of the selected metabolites, as they
appear to display specific properties. 7-Hydroxymethotrexate, which
pharmacological activity is controversial (Fabre and Goldman, 1985;
 J.-P. Besse et al. / Environment International 39 (2012) 73–86 79

Table 4
Preferential list of anticancer drugs for the aquatic environment, based on calculated PECs, available pharmacological, metabolism and biodegradation data.

Molecule Mechanism PEC Excretion WWTP Refined Conclusion Discussion

of action (ng.l− 1) fraction removal rate/ PEC (ng.l− 1)
Biodegradation

Hydroxycarbamide Other 156.13 0.5 nd 78.1 Preferential 1) Hydroxycarbamide (Hydroxyurea) is by far

the most consumed cytotoxic molecule (44% and 39%
of the total amounts consumed in 2004 and 2008 respectively)
2) Need to confirm its occurrence in wastewaters
and surface waters and if necessary to conduct
ecotoxicological assays
Capecitabine Antimetabolite 117.24 0.03 0.15a 2.9 Preferential 1) Precursor of Fluorouracil
2) Despite a very low excretion fraction, the refined
PEC is still higher than 1 ng.l− 1
3) Even if the highest biodegradation reported value is used,
the refined PEC is higher than 1 ng.l− 1
4) Need to assess occurrence of this molecule
in the environment
Fluorouracil Antimetabolite 39.57 0.2b >0.9c b0.8 Non 1) No need to assess for its occurrence considering
preferential current scientific knowledge
Imatinib Tk inhibitor 19.95 0.25 nd 5 Preferential 1) No data available on environmental behavior
2) Need to confirm its presence in the aquatic environment
d
Bicalutamide Antiandrogen 19.7 0.55 not readily 15.7 Preferential 1) Possibly only weakly active if (S)-enantiomer is the main
biodegradablee excreted compound (see reference d)
2) Need to confirm its presence in the aquatic environment
Flutamide Antiandrogen 11.9 b0.1 nd b1.2 Non 1) No need to assess for occurrence considering
preferential current scientific knowledge
Gemcitabine Antimetabolite 8.66 0.1 0.4f 0.52 Non 1) No need to assess for occurrence considering
preferential current scientific knowledge
Tamoxifen Antiestrogen 8.61 nd Not readily nd Preferential 1) Only limited data on excretion
biodegradableg 2) Could be excreted in significant amounts in feces
as unchanged drug or glucuronide
(BCB, 2011; Kisanga et al., 2005)
3) Already detected in surface waters and wastewaters
Cyclophosphamide Alkylating 6.98 >0.25 0h >1.75 Preferential 1) Expected to occur in the environment at the
agent ng.l− 1 range due to its very low biodegradability
(Kümmerer et al., 2000)
2) Refined PECs are consistent with field measurements
Estramustine Alkylating 6.57 nd nd nd nd 1) Only approximately 75% of the drug is orally absorbed, and
agent unchanged Estramustine can be excreted through biliary route
2) This molecule might be present in the aquatic environment
at concentrations higher than the ng.l− 1, but metabolism
and degradation data are too scarce to allow drawing any
conclusion.
Mitotane Other 5.34 0.6 nd 3.2 Preferential 1) No data available on environmental behavior
2) Need to confirm its presence in the aquatic environment
Exemestane Antiaromatase 4.16 nd >0.9i b0.4 Non 1) No need to assess for occurrence considering current
preferential scientific knowledge
Nilutamide Antiandrogen 3.86 b0.02 nd 0.07 Non 1) No need to assess for occurrence considering current
preferential scientific knowledge
Erlotinib Tk inhibitor 3.4 b0.02 Not readily b0.07 Non 1) No need to assess for occurrence considering current
biodegradablej preferential scientific knowledge
Cytarabine Antimetabolite 3.05 b0.1 0.5k b0.15 Non 1) No need to assess for occurrence considering current
preferential scientific knowledge
Lapatinib Tk inhibitor 2.65 0.7l nd 1.82 Preferential 1) The excretion rate used here is the maximum reported
value and might not be representative of the general case.
2) If median excretion value is used (Appendix A),
Lapatinib is no longer a preferential compound
Ifosfamide Nitrogen 2.35 0.5 0m 1.18 Preferential 1) Expected to occur in the environment at the
mustard ng.l− 1 range due to its very low biodegradability
(Kümmerer et al., 1997, 2000)
2) Refined PECs are consistent with field measurements
Mercaptopurine Antimetabolite 2.17 0.07 nd 0.15 Non 1) No need to assess for its occurrence considering
preferential current scientific knowledge
Bevacizumab mab 1.99 nd nd nd nd 1) Data too scarce to conclude
Carboplatin Platinum 1.91 1 0.65n 0.67 Non 1) Reported to be removed in wastewater and
derivative preferential during water treatment processes,
2) Adsorption to sludge and solid particles
(Lenz et al., 2005, 2007)
Methotrexate Antimetabolite 1.71 0.9 0.95o 0.07 Non 1) No need to assess for its occurrence considering
preferential current scientific knowledge
2) Refined PEC could be higher as Methotrexate
is also used for the treatment of Rheumatoid
polyarthritis under another ATC class (L04)
Rituximab mab 1.65 nd nd nd nd 1) Data too scarce to conclude
Pipobroman Alkylating 1.53 nd nd nd nd 1) Data too scarce to conclude
agent

(continued on next page)

80 J.-P. Besse et al. / Environment International 39 (2012) 73–86

Table 4 (continued)
Molecule Mechanism PEC Excretion WWTP Refined Conclusion Discussion
of action (ng.l− 1) fraction removal rate/ PEC (ng.l− 1)
Biodegradation

Nilotinib Tk inhibitor 1.34 0.6 nd 0.8 Non 1) No need to assess for occurrence considering
preferential current scientific knowledge
Trastuzumab mab 1.28 nd nd nd nd 1) Data too scarce to conclude
Cetuximab mab 1.26 nd nd nd nd 1) Data too scarce to conclude
Temozolomide Alkylating 1.22 0.1 nd 0.12 Non 1) No need to assess for occurrence considering current
agent preferential scientific knowledge
Irinotecan Other 1.06 >0.5 nd >0.5 nd 1) Data too scarce to conclude

nd: not determined.

Tk inhibitor: tyrosine kinase inhibitor
mab: monoclonal antibody.
a
OECD 302B, 15% removal at 7 days, up to 58% removal at 28 days (reported values from Straub, 2010).
b
Taking into account the fraction of Capecitabine excreted as Fluorouracil.
c
Batch experiment, incubated with activated sludge (Mahnik et al., 2004); complete biodegradation with test OECD 303A (Kiffmeyer et al., 1998); reported not biodegradable in
the closed bottle test and Zahn–Wellens test (Kümmerer and Al-Ahmad, 1997).
d
Data from FASS, 2011. Bicalutamide is a racemate, which activity resides almost exclusively in the (R)-enantiomer, and a part of the excreted Bicalutamide is the inactive
(S)-enantiomer (Cockshott, 2004).
e
Not readily biodegraded: OECD 301C, OECD 302A (FASS, 2011).
f
Closed bottle test and Zahn–Wellens test, from 42% to 50%, depending on tests conditions (Kümmerer and Al-Ahmad, 1997).
g
Not readily biodegraded: OECD 301C (FASS, 2011).
h
Suggested to be persistent in WWTP (Buerge et al., 2006); reported not biodegradable in closed bottle test and Zahn–Wellens test (Kümmerer et al., 2000).
i
Not considered to be persistent, OECD 308 (Ericson, 2007).
j
Not readily biodegraded, OECD 301C (FASS, 2011).
k
Closed bottle test and Zahn–Wellens test, from 50% to more than 95%, depending on test conditions (Kümmerer and Al-Ahmad, 1997).
l
Considering maximal excretion rate (median excretion rate of 27%, see Appendix A for details).
m
Reported not biodegradable in closed bottle test and Zahn–Wellens test (Kümmerer et al., 2000).
n
From 65% to 82% removal rate with membrane bioreactor alone or with advanced sewage treatment processes (Lenz et al., 2007).
o
OECD confirmatory test (Kiffmeyer et al., 1998); reported as readily biodegradable with OECD 301F test (Henschel et al., 1997).

Lingg et al., 2005; Seither et al., 1989), has been reported to appear
during biodegradation experiments of Methotrexate, and further-
more, not biodegradable (Kiffmeyer et al., 1998). It remains unclear
to which extent the biodegradation of Methotrexate can lead to the
formation of 7-hydroxymethotrexate. Nevertheless, as this metabolite
is potentially active, not readily biodegradable, and as it could be
found in the aquatic environment at levels higher than the parent
compound, occurrence studies should be conducted.
The antiestrogenic Tamoxifen is metabolized into two active
metabolites, Hydroxy-Desmethyltamoxifen (Endoxifen) and 4-
Hydroxytamoxifen. The two metabolites have been reported to be
100 times more potent and active than the parent compound and
are considered responsible for the pharmacological activity (BCB,
2011; Micromedex Drugdex©, 2011; Wu et al., 2009; Zheng et al.,
2007). Available metabolism data do not allow determining a reliable
PEC for these metabolites, but they have been detected in bile and
urine (Kisanga et al., 2005). Considering their high antiestrogenic activity
and their potential presence in the aquatic environment, these metabo-
lites should be investigated for occurrence, and if detected in the aquatic
environment, tested for ecotoxicity, as they might be of greater concern
than the parent Tamoxifen.
Searching for metabolites in surface waters can be costly and hard to
achieve due to limited available standards, and because they are expected
to be present at low concentrations. Nevertheless, pharmaceutical metab-
olites, which have already been detected in surface waters and sediment
(Langford and Thomas, 2010), and subsequently anticancer metabolites,
require consideration as they can exert biological effects on aquatic organ-
isms. Moreover, they should be taken into account when assessing the
risk of pharmaceuticals (Besse et al., 2008; Celiz et al., 2009; EMEA,
2006). To this extent, the work provided here is incomplete as only
human metabolites were considered, but no environmental degradation
compounds such as photolysis by-products. Nevertheless, current knowl-
edge is probably insufficient to support any conclusion on the latter.
5.2. Implications of consumption trends
National and local consumption data both show an increase in the
consumption of anticancer drugs. As consumption trends directly
influence levels of anticancers entering the aquatic environment,
it therefore appears necessary to continue surveying their presence
in WWTP effluents and surface waters.
Consumption data also show that hospital effluents may not be the
primary input source of anticancer drugs in the environment. Indeed,
a distinct shift in consumption trends has been observed between
2004 and 2008, which appears to be linked with the generalization
of home treatment in order to improve patients' comfort. The fact
that an increasing number of molecules originally restricted to hospitals
are now available in town pharmacies, combined with the development
of such home treatments, leads to a more diffuse rejection of anticancers
throughout a given town. This contradicts the generally admitted idea
that hospital effluents are the primary input source of anticancer drugs
in the environment.
Local consumption data from the Centre Léon Berard (CLB), in
Lyon, which allows monitoring the distribution patterns of anticancer
drugs depending on the type of hospitalization (complete hospitalization
and outpatients), provide a meaningful illustration for this. The hospital-
ization type strongly influences the entrance pathway of anticancer
drugs in the sewage system. Molecules consumed by patients during
complete hospitalization at the CLB are 100% evacuated with the center's
effluents. On the contrary, only up to 20% of the drugs delivered to
outpatients are considered consumed on site. Therefore, only 20% of
outpatient delivered drugs are eliminated with hospital effluents, while
the remaining 80% are consumed by patients at home and thus are
discharged into the sewer system diffusely throughout the town and
its suburbs. In 2004 and 2008, amounts delivered to outpatients
accounted for approximately half of the total amounts consumed at
the CLB. Based on available AFSSAPS and CLB data, we established a the-
oretical scheme of anticancer drugs inputs in the aquatic environment
(Fig. 3). These results are of importance for anticancer drug risk
management.
Data available here seem to indicate that this shift in consumption
trends is observed both at a national and a local scale. However, such an
assumption should be confirmed with other recent national and local
data sources, originating from other hospitals and cancer care centers.
Finally, if hospital effluents are no longer the main expected entry
route of anticancer drugs in the environment, they nevertheless
 J.-P. Besse et al. / Environment International 39 (2012) 73–86 81

Table 5
Preferential list of anticancer metabolites for the aquatic environment, based on calculated PECs, available pharmacological, metabolism, and biodegradation data.

Molecule Active Parent compound Excretion PEC Conclusion Discussion

(PEC, ng.l− 1) fraction (ng.l− 1)

N-desmethylmetabolite Yes Imatinib (19.95) 0.13 2.6 Preferential 1) No degradation data available
(CGP74588) 2) Need to confirm occurrence in waters
Hydroxy-bicalutamide No Bicalutamide 0.32 6.3 Preferential 1) No degradation data available
(19.7) 2) Need to confirm occurrence in waters
Hydroxy-flutamide Yes Flutamide (11.5) nd nd nd 1) Data too scarce to conclude
2′,2′-Difluorodeoxyuridine Yes Gemcitabine 0.9 7.8 Preferential 1) Reported to be pharmacologically active
(dFdU) (8.66) (Benyumov et al., 2011; Veltkamp et al., 2008)
2) Suspected to contribute to the toxicity of Gemcitabine
(Micromedex Drugdex©, 2011)
3) Occurrence and, if necessary, ecotoxicity studies should be conducted
4-Hydroxy-N- Yes Tamoxifen (8.61) nd nd Preferential NB: otherwise indicated, the following considerations concern the two
desmethyltamoxifen metabolites of Tamoxifen:
(Endoxifen) 1) 100 times more affinity to estrogen receptor than Tamoxifen
(BCB, 2011; Wu et al., 2009; Zheng et al., 2007)
2) 100 times more potent than Tamoxifen
(Wu et al., 2009; Zheng et al., 2007)
3) Could be excreted as glucuronide conjugates
4) Detected in urine, but no excretion rates are available
(Kisanga et al., 2005)
5) Hydroxy-tamoxifen could be the main metabolite of Tamoxifen
6) No degradation data are available
7) Occurrence is to be confirmed and if, necessary, ecotoxicity studies should
be conducted for these two metabolites
4-Hydroxy-tamoxifen Yes Tamoxifen (8.61) nd nd Preferential
Phosphoramide mustard Yes Cyclophosphamide 0.4 b1 Non 1) Considered responsible for pharmacological effects
(6.98) preferential 2) Reported to be rapidly hydrolyzed: half-life at pH 7.4
and ambient temperature of 2 h (Joqueviel et al., 1998)
3) No need to assess for its occurrence considering current scientific
knowledge
Estromustine Yes Estramustine nd nd nd 1) Reported to be cytotoxic (BCB, 2011)
(6.57) 2) Neither excretion nor biodegradation data are available
3) Not possible to conclude on its occurrence in the environment
17-hydro-exemestane Yes Exemestane (4.16) nd nd nd 1) Data are too scarce to conclude
N-Desmethylmetabolite Yes Erlotinib (3.4) 0.03 0.1 Non 1) No need to assess for occurrence, considering
(OSI-420) preferential current scientific knowledge
Carboxylic acid metabolite nd Erlotinib (3.4) 0.3 1 Preferential 1) No degradation data available
2) Need to confirm occurrence in water
Uracil-arabinoside No Cytarabine (3.05) 0.9 2.74 Preferential 1) No degradation data available
2) Need to confirm occurrence in water
7-Hydroxy-methotrexate Yes Methotrexate 0.15 nd Preferential 1) Pharmacological activity unclear (see Section 5.1.2)
(1.71) 2) Reported to appear during biodegradation
experiment of Methotrexate, (Kiffmeyer et al., 1998)
3) Reported as not biodegradable (Kiffmeyer et al., 1998)

nd: not determined.

remain a specific input source for specific pharmaceuticals (i.e. re-
stricted to hospitals) and other macro and micropollutants (Verlicchi
et al., 2010), or even pathogenic microorganisms (Emmanuel et al.,
2009; Prado et al., 2011).
Overall, variability and trends in consumption amounts illustrate
the need to make consumption data public and easily accessible for
anticancer drugs but also for all other pharmaceutical products. Public
access of consumption amounts allows determining what is currently
entering the aquatic environment and what are the input pathways.
Subsequent exposure assessments, such as the one implemented
here, allow presenting an exhaustive overview of molecules likely to
be of environmental concern and are helpful for any further occurrence
or ecotoxicological investigation. Furthermore, they are the basis of any
comprehensive risk assessment.
5.3. Risk considerations
5.3.1. Cytotoxic compounds
Currently, only very few ecotoxicological data are available for
cytotoxic drugs (Table 6a). Recently, an extensive risk assessment
based on the EMEA guideline (EMEA, 2006) was conducted for
Capecitabine and Fluorouracil (Straub, 2010). On the basis of PECs,
available measured concentrations, available ecotoxicological data,
and new chronic data, built using standardized chronic assays, it
was concluded that there was no exposure level of concern in the
aquatic environment for these molecules.
Furthermore, some chronic data, based on standardized chronic
tests on Daphnia magna, suggest that Cytarabine and Gemcitabine are
also of limited environmental concern at their current environmental
exposure level (Zounkova et al., 2010).
As quoted by several authors (Johnson et al., 2008; Kümmerer and
Al-Ahmad, 1997), genotoxic effects of cytotoxic drugs are expected to
be a major issue for the aquatic environment. Studies conducted until
now, and based on in vitro assays, suggest that the environmental
concentrations of tested cytotoxics are under concentrations required
to cause genotoxic effects (Ferk et al., 2009; Zounkova et al., 2010, 2007).
Overall, it appears that most of the molecules have been tested only
by acute standardized tests or genotoxic in vitro assays (Table 6a), mak-
ing it difficult to conclude on their environmental risk. Therefore, stud-
ies need to be implemented in order to acquire relevant
ecotoxicological data. To this extent, the suitability of standardized
tests and endpoints for the assessment of cytotoxic drugs is arguable:
if acute tests measuring survival endpoints are clearly inadequate, as
shown on other pharmaceutical classes (Kümmerer et al., 2004), the
use of algal growth, Daphnia magna reproduction, and even fish sub-
chronic tests can be questioned, as they do not seem suitable for
82 J.-P. Besse et al. / Environment International 39 (2012) 73–86

Total delivered amounts or surface waters. However, since they only measure subcellular
effects, it is difficult to relate observed effects on a single compound
78% of the 22% of the (or even a mixture) with potential environmental effects at the indi-
total amountsa total amountsa vidual or population level. We therefore consider that no conclusion
Town Hospitals
on environmental genotoxic effects should be drawn on the sole
basis of in vitro assays. The in vivo comet assay, which is increasingly
10.3%b 11.7%b used to assess genotoxicity in organisms exposed to contaminants
8.2%c in situ or in laboratory experiments (Frenzilli et al., 2009; Lacaze
Consumption at Complete
Outpatients et al., 2010), could be a more suitable tool to investigate the geno-
home hospitalization
toxic effects of cytotoxic compounds on aquatic species. A possible per-
2.1%d 11.7% spective could be the assessment of genotoxic effects on spermatozoa
and oocytes of exposed aquatic organisms (Lacaze et al., 2011). Germi-
Diffuse rejection
in sewer system
Hospital effluents nal DNA cell damage, related to poor fertilization rate and defective
embryonic development (Aitken and De Iuliis, 2007), can lead to re-
86.2% 13.8% production impairment, which is a relevant ecotoxicological endpoint
for anticancer drugs at the individual, population, and community scale.
urban WWTP

Fig. 3. Theoretical input pathways for anticancer drugs in the aquatic environment
with relative contribution of each entry pathway. Based on gathered French data
from the AFSSAPS (national consumption amounts) and from the Centre Léon Bérard
(CLB; specialized cancer care center); see Section 4.2 and Appendix B for details.
a
: Using AFSSAPS data (AFSSAPS, 2009a). b: Using CLB data, approximately 53% of the
total amounts were delivered for complete hospitalization and 47% for outpatients, at
the CLB in 2008 (Appendix C). c: Approximately 80% of the anticancer drugs delivered
to outpatients are considered consumed and rejected at home (i.e. outside the CLB).
d
: Approximately 20% of the anticancer drugs delivered to outpatients are considered
consumed and rejected onsite.
determining long-term effects of exposure to low concentrations of
DNA interacting molecules. To our knowledge, no experiments have
been conducted on long-term toxicity of cytotoxic drugs in fish, mol-
luscs, or invertebrates other than Daphnia magna. Even if hard to imple-
ment, full-life cycle or multigenerational assays on fish sound more
relevant for assessing environmental chronic effects of cytotoxics.
Moreover, as cytotoxics may contribute to the genotoxicity
detected in certain effluent or surface water samples (Jolibois and
Guerbet, 2005; Reifferscheid and Grummt, 2000), their genotoxic
potential should be tested. To do so, we consider that in vivo assays
should be more suitable than in vitro tests. Indeed, in vitro assays
are useful tools for screening genotoxic effects, notably in effluents
5.3.2. Cytostatic compounds
Cytostatics are a recent class of anticancer drugs considered as
promising for the improvement of treatment efficiency. Two sub-
groups can be found among this class: Protein kinase inhibitors (in
particular tyrosine kinase inhibitors) and monoclonal antibodies.
Tyrosine kinases belong to the protein kinase enzyme family. Protein
kinases are key regulators of cell function that constitute one of the largest
and most functionally diverse gene families. By adding phosphate groups
to substrate proteins, they direct the activity, localization and overall
function of many proteins. Protein kinases therefore mediate most of
the signal transduction in eukaryotic cells, regulating cellular metabolism,
transcription, cell cycle progression, cytoskeletal rearrangement and cell
movement, apoptosis, and differentiation. The diversity of essential
functions mediated by kinases is shown by the conservation of some
50 distinct kinase families between yeast, invertebrate and mammalian
kinomes (Manning et al., 2002). Currently, nothing is known on their
occurrence and biological effects (if any) on aquatic non-target organ-
isms, and only limited data are available for Erlotinib (Table 6a). Never-
theless, considering their universal mode of action and increasing
consumption amounts, they should be taken into consideration and
assessed for occurrence, and if need be, for ecotoxicity.

Table 6a
Review of existing ecotoxicological data for anticancer drugs (cytotoxic and cytostatic molecules). Concentrations are expressed in mg.l− 1.

Molecule Organism Endpoint Value Concentration (mg.l− 1) References

Methotrexate V.fischerii Luminescence EC50 1220 Henschel et al. (1997)

S.subspicatus Growth EC50 72 h 260 Henschel et al. (1997)
T. pyriformis Growth EC50 48 h 45 Henschel et al. (1997)
D. magna Immobilization EC50 48 h >1000 Henschel et al. (1997)
B. rerio Survival EC50 96 h 85 Henschel et al. (1997)
X. laevis Growth EC50 96 h 0.015 Bantle et al. (1994)
Fluorouracil V. fischerii Luminescence EC50 0.12 Backhaus et al. 2000
P. promelas Growth LOEC 120 h 20 De young et al. (1996)
D. magna Reproduction LOEC 21 days 0.05 Zounkova et al. (2010)
D. magna Reproduction NOEC 21 days 0.0028 Straub (2010)
A. flos-aquae Growth NOEC 72 h 0.002 Straub (2010)
Capecitabine D. magna Reproduction EC50 48 h >850 Straub (2010)
P. subcapitata Growth NOEC 72 h 0.14 Straub (2010)
Cladribin D. magna Immobilization EC50 48 h 233 FDA-CDER (1996)
Cytarabine D. magna Reproduction LOEC 21 days 3.7 Zounkova et al. (2010)
Gemcitabin D. magna Reproduction LOEC 21 days >1.0 Zounkova et al. (2010)
P. subcapitata Growth EC50 72 h 0.57 FASS (2011)
D.magna Immobilization EC50 48 h >0.99 FASS (2011)
P. promelas Survival LC50 96 h >1000 FASS (2011)
O. mykiss Survival LC50 96 h >1000 FASS (2011)
Paclitaxel D. magna Immobilization EC50 48 h >0.74 FDA-CDER (1996)
Thiotepa D. magna Immobilization EC50 48 h 546 FDA-CDER (1996)
Erlotinib S. capricornutum Growth NOEC 72 h 0.14 FASS (2011)
D. magna Reproduction NOEC 48 h 0.7 FASS (2011)
O. mykiss Survival NOEC 14 days 0.02 FASS (2011)
 J.-P. Besse et al. / Environment International 39 (2012) 73–86 83

Monoclonal antibodies are molecules that target growth factor
receptors in order to exert a direct effect on the growth and survival
of cancer cells by antagonizing ligand-receptor signaling. Monoclonal
antibodies can target to cell surface antigens and directly elicit apoptotic
signaling (Ludwig et al., 2003). Conservative PEC calculated here are
very low for this class of compound. Considering available knowledge,
it is currently impossible to conclude on the environmental significance
of this class of molecules; again, there is a need for further scientific
investigation.
5.3.3. Endocrine therapy drugs
Due to their mechanism of action, exposure to such molecules
could be hazardous for aquatic species, notably fishes, batracians or
molluscs, displaying steroid receptors similar to mammals.
5.3.3.3. Antiestrogens. Among the molecules investigated, Tamoxifen is
the only one for which some occurrence and ecotoxicological data are
available (Table 6b). A full-life cycle study has been conducted on
Pimephales promelas (Williams et al., 2007). Several endpoints were
measured and an overall NOEC for F0 and F1 generations was
reported to be equal to 5.12 μg.l − 1, which is far above expected con-
centrations in water. Authors therefore considered that Tamoxifen is
not expected to pose a risk at the current environmental levels of
exposure. However, since Tamoxifen is metabolized into two more
potent metabolites, potentially present in the aquatic environment,
the latter should be taken into account when assessing Tamoxifen
ecotoxicological effects and environmental risk.
No data are available for Toremifen, however considering its very low
conservative PEC (about 0.02 ng.l− l), risk for the aquatic environment is
expected to be limited.

5.3.3.1. Hormone analogs. Chronic administration of these molecules

in human results in a significant reduction of plasma testosterone.
Considering this, occurrence of these molecules in the aquatic envi-
ronment may be harmful for sensitive non target-organisms. However,
given their very low consumption amounts and very low PEC values
(lower than 0.1 ng.l− 1, see Table 2), the associated risk is expected to
be limited relative to other endocrine disruptors, even if data remain
too scarce to draw any definitive conclusion.
5.3.3.2. Anti-aromatase molecules. These molecules act by inhibiting
the aromatase activity, which converts androgens into estrogens. All
compounds have very low PECs, either conservative or refined
(Tables 2 and 4). Very few ecotoxicological data are available for these
molecules (Table 6b). Available data for Letrozole (Sun et al., 2007)
and some data available for Fadrozole (Gust et al., 2010; Panter et al.,
2004), an anti-aromatase molecule no more used in human medicine,
report effects at the μg.l− 1 level. Studies comparing the in vitro inhibitory
aromatase activity of different molecules (Trösken et al., 2004, 2006)
show that Letrozole is the most potent aromatase inhibitor among
those tested, with an inhibitory activity stronger than Procholoraz,
Flusilazole and Imazalil (molecules used as fungicides in agriculture).
Considering their mechanism of action and antiaromatase potency,
these molecules could lead to a disruption of reproductive functions in
non-target aquatic organisms, especially molluscs and fishes. However,
given the very low PEC values for compounds investigated here, and
given the lack of ecotoxicological data, it is not possible to conclude
for these molecules.
5.3.3.4. Antiandrogens. To our knowledge, no occurrence data for the
aquatic environment are available, and only a few ecotoxicological
studies have been conducted on these molecules (Table 6b). Calculated
PECs for these molecules are low, with the higher PEC observed for Bica-
lutamide, which is notably reported as not biodegradable (Table 4). Con-
sidering PEC values, risk posed by Nilutamide and Fulvestrant is expected
to be limited relative to other compounds from the same class. Ecotoxico-
logical data on fish indicate toxic concentrations far above PEC for Fluta-
mide, Nilutamide and even Bicalutamide (Table 6b). Overall,
considering PECs calculated here and limited available ecotoxicological
data, risk is expected to be limited relative to other endocrine dis-
rupters, but further works are needed prior to conclude definitely.
5.3.3.5. 5-α-reductase inhibitors. Finasteride and Dutasteride are 5-α-
reductase inhibitors that inhibit the conversion of testosterone to
dihydrotestosterone. These molecules are not strictly speaking anticancer
drugs but considering their mechanism of action and indications (i.e.
treatment of prostatic hyperplasia), we discuss them in this paper. No
occurrence studies have been conducted for these two molecules in the
aquatic environment. As gathered consumption amounts were incom-
plete (Finasteride is also used in the treatment of androgenic alopecia),
no reliable PEC could be calculated. Nevertheless, available consumption
data indicate that such a PEC is expected to be in the ng.l− 1 range.
Moreover, Finasteride consumption has increased since 2004.
Only limited ecotoxicological data are available for Finasteride; which
report impairment of gonad development and sexual differentiation in

Table 6b
Review of existing ecotoxicological data for anticancer drugs (endocrine therapy drugs). Concentrations are expressed in μg.l− 1.

Molecule Organism Endpoint Value Concentration (μg.l− 1) References

Letrozole O. latipes Fecundity LOEC 21 days 5 Sun et al. (2007)

O. latipes Fertility LOEC 21 days 5 Sun et al. (2007)
O. latipes Increase in genotypic F1 males LOEC 21 days 5 Sun et al. (2007)
Tamoxifen P. promelas Overall a NOEC 284 days 5 Williams et al. (2007)
P. promelas F1 larvae growth b 28 days 0.08 Williams et al. (2007)
P. promelas F1 growth c 112 days 0.01 Williams et al. (2007)
P. promelas Increase in Vtg, F1 males d 112 days 0.01 Williams et al. (2007)
A. tonsa Larval development EC50 5 days 49,000 Andersen et al. (2001)
Nilutamide Green algae Growth NOEC 1000 FDA-CDER (1996)
Flutamide B. calicyflorus Fertilization of sexual females LOEC 96 h 1 Preston et al. (2000)
G. aculeatus Spiggin inhibition LOEC 21 days 500 Sebire et al. (2008)
G. aculeatus Male behavior LOEC 21 days 100 Sebire et al. (2008)
P. promelas Testis alterations LOEC 21 days 62 Jensen et al. (2004)
P. promelas Increase of estradiol plasma levels LOEC 21 days 651 Jensen et al. (2004)
Bicalutamide P. promelas Overall e NOEC f 10 FASS (2011)
a
NOEC value considered significant for all endpoints in F0 and F1 generations.
b
Statistically significant decrease in F1 larvae length and weight at 28 days post hatch.
c
Statistically significant decrease in F1 individuals length and weight at 112 days post hatch, observed at this concentration only.
d
Statistically significant increase in vitellogenin in F1 males at 112 days post hatch, observed at this concentration only.
e
Overall endpoints : reproduction, weight, length, secondary sexual characteristics, time to hatch, hatchability, survival and histology.
84 J.-P. Besse et al. / Environment International 39 (2012) 73–86
male tadpoles at the mg.l− 1 range (Urbatzka et al., 2009). As for other
antiandrogens, data remain too scarce to make any conclusion on
these two molecules.
5.4. Assessing the environmental effects of anticancer drugs
As quoted by other authors (Johnson et al., 2008), we consider
that it is still not possible to draw any definitive conclusions on
risks posed by these molecules. Regarding EMEA guidelines, risks
associated to anticancer drugs are generally discussed relative to
isolated molecules, and so are subjected to ecotoxicological assays.
Nevertheless, the true question is: what are the effects related to a
chronic exposure to a mixture of several molecules with similar mech-
anisms of action, each at very low concentrations in the ng.l − 1 range
(or even less)? Biotests on isolated molecules, relying on standardized
endpoints, provide meaningless information in the case or anticancer
drugs. Although currently necessary for regulatory considerations, we
question the suitability of these tests for assessing environmental ef-
fects of such compounds.
Due to their mechanisms of action and very low expected environ-
mental concentrations, anticancer drugs require taking a step forward
from “classic” standardized approaches. It is our opinion that there is a
need to better account for the environmental reality and complexity by
testing mixtures of anticancers at environmentally realistic concentra-
tions (i.e. at the ng.l− 1 range). Testing for mixtures is still an emerging
and complicated field but existing work shows how important this
issue is (Brian et al., 2007; Sárria et al., 2011; Sun et al., 2009).
Mixtures tests for anticancer drugs could be implemented according
to i) an “environmental” approach, taking into account molecules likely
to be present in the environment; and ii) a “pharmacological” approach
based on the combination of molecules used in cancer treatment.
Chemotherapy protocols are often based on combinations of several
molecules with different targets, kinetics, and mechanisms of action, in
order to obtain additive or even synergistic effects and to enhance ther-
apeutic effects. For example, for the treatment of hormone-dependant
breast cancers, chemotherapy based on cytotoxic compounds is used
in association with hormonotherapy. Furthermore, the combined use
of Methotrexate, Rituximab and Temozolomide is applied for the
treatment of certain lymphomas. Fluorouracil and Imatinib are a combi-
nation proposed for the treatment of refractory pancreatic cancers.
There are plenty of other examples of such therapeutic combinations.
In our opinion, such an approach could allow testing for relevant
mixtures, with a potential environmental significance.
This is the same for antiestrogens, where the risk of mixture with
molecules displaying the same mechanism of action seems to be the
major point to consider. Despite it is not used for cancer therapy (it
belongs to another ATC class), Raloxifene, used in osteoporosis, displays
the same mechanism of action than Tamoxifen. Raloxifene has a refined
PEC of 71 ng.l− l (data not shown). Therefore, rather than investigating
the ecotoxicity of isolated Tamoxifen, it seems sound to evaluate the
cumulative effects of Tamoxifen, Raloxifene and Tamoxifen metabolites,
especially on aquatic organisms displaying steroid receptors structurally
close to mammals (e.g. fishes, batracians or molluscs).
6. Conclusion
Despite an extensive review on existing consumption and on
pharmacological and ecotoxicological data, results of this study are
somewhat frustrating, as it appears that very little is known about the
environmental significance of anticancer drugs. Currently, the scientific
knowledge on the presence and effects of anticancer drugs in the aquatic
environment remains insufficient to support a definitive conclusion. If
one can speculate on the environmental effects of cytotoxic and endo-
crine therapy drugs, the environmental effects (if any) of tyrosine kinase
inhibitors and monoclonal antibodies remain unknown, and studies
should be implemented on these classes of molecules.
It is our point of view that many steps remain to be taken concerning
the environmental impact of anticancer drugs, and that assessing the
environmental risk they pose is a challenging though interesting and
necessary issue for the coming years.
To summarize our opinion on anticancer drugs, there is a need to:
• assess the occurrence of parent anticancer drugs and metabolites
highlighted here;
• break away from “classical” risk assessment based on standardized
tests;
• assess the effect of environmental mixtures of anticancer drugs at
environmentally realistic concentrations;
• focus on in vivo long term, even full life cycle tests, for the assessment
of ecotoxicity, genotoxicity and/or endocrine disrupting effects of
anticancer drugs.
Acknowledgments
The authors wish to thank the AFSSAPS and particularly Paul
Houeto, for kindly sharing the pharmaceutical consumption data.
They also appreciate Pr François Locher, from the Faculty of Pharmacy of
Lyon, and Monique Boucquin, from the “Pharmaceutical Documentation
and Information Center (CDIP)” of Lyon hospices, for providing pharma-
ceutical information. Finally the authors are very grateful to Ashley
Tilghman-Sibille for proofreading and correcting the English, and also to
the two anonymous reviewers, for their helpful comments that allowed
us improving this article.
Supplementary data (Appendices A, B and C)
Supplementary data to this article can be found online at doi:10.
1016/j.envint.2011.10.002.
References
AFSSAPS. Agence Française de Sécurité Sanitaire des Produits de Santé. French con-
sumption amounts for pharmaceuticals, including anticancer drugs (ATC class
L01) for year 2004. 2006. Personal communication.
AFSSAPS. Agence Française de Sécurité Sanitaire des Produits de Santé. French con-
sumption amounts for anticancer drugs (ATC class L01) for year 2008. 2009a. Per-
sonal communication.
AFSSAPS. Agence Française de Sécurité Sanitaire des Produits de Santé. French con-
sumption amounts for endocrine therapy drugs (ATC class L02) for year 2008.
2009b. Personal communication.
Aherne GW, Hardcastle A, Nield AH. Cytotoxic drugs and the aquatic environment. Es-
timation of Bleomycin in river and water samples. J Pharm Pharmacol 1990;42
(10):741–2.
Aitken RJ, De Iuliis GN. Origins and consequences of DNA damage in male germ cells.
2007. Reprod Biomed Online 2007;14(6):727–33.
Andersen HR, Wollenberger L, Halling-Sorensen B, Kusk KE. Development of copepod
nauplii to copepodites. A parameter for chronic toxicity including endocrine dis-
ruption. Environ Toxicol Chem 2001;20(12):2821–9.
Ashton D, Hilton M, Thomas KV. Investigating the environmental transport of human
pharmaceuticals to streams in the United Kingdom. Sci Total Environ 2004;333:
167–84.
Backhaus T, Altenburger R, Boedeker W, Faust M, Scholze M, Grimme LH. Predictability
of the toxicity of a multiple mixture of dissimilarly acting chemicals to Vibrio
fischeri. Environ Tox Chem 2000;19(9):2348–56.
Bantle JA, Burton DT, Dawson DA, Dumont JN, Finch RA, Fort DJ, et al. Rayburn JR Fetax
interlaboratory validation study: phase II testing. Environ Toxicol Chem 1994;13
(10):1629–37.
BCB. Banque Claude Bernard. http://www.resip.fr. 2011. Last accessed august 2011.
Benyumov A, Gurvich VJ, Lis LG, Williams BW, Kirstein MN. Combinatorial pharmaco-
logic effects of gemcitabine and its metabolite dFdU. ChemMedChem 2011;6(3):
457–64.
Besse J-P, Garric J. Human pharmaceuticals in surface waters. Implementation of a pri-
oritization methodology and application to the French situation. Toxicol Lett
2008;176(2):104–23.
Besse J-P, Garric J. Progestagens for human use, exposure and hazard assessment for
the aquatic environment. Environ Pollut 2009;157(12):3485–94.
Besse J-P, Kausch-Barreto C, Garric J. Exposure assessment of pharmaceuticals and their
metabolites in the aquatic environment: application to the French situation and
preliminary prioritization. Hum Ecol Risk Assess 2008;14(4):665–95.
Brian JV, Harris CA, Scholze M, Kortenkamp A, Booy P, Lamoree M, et al. Evidence of es-
trogenic mixture effects on the reproductive performance of fish. Environ Sci Tech-
nol 2007;41(1):337–44.
 J.-P. Besse et al. / Environment International 39 (2012) 73–86
Buerge IJ, Buser H-R, Poiger T, Müller MD. Occurrence and fate of the cytostatic drugs
cyclophosphamide and ifosfamide in wastewater and surface waters. Environ Sci
Technol 2006;40(23):7242–50.
Catastini C, Mullot J-U, Boukari S, Mazellier P, Levi Y, Cervantes P, et al. Assessment of
antineoplastic drugs in effluents of two hospitals | [Identification de molécules
anticancéreuses dans les effluents hospitaliers]. J Eur Hydrol 2008;39(2):171–80.
Celiz MD, Tso J, Aga DS. Pharmaceutical metabolites in the environment: analytical
challenges and ecological risks. Environ Toxicol Chem 2009;28(12):2473–84.
Cockshott ID. Bicalutamide: clinical pharmacokinetics and metabolism. Clin Pharmaco-
kinet 2004;43(13):855–78.
Coetsier CM, Spinelli S, Lin L, Roig B, Touraud E. Discharge of pharmaceutical products
(PPs) through a conventional biological sewage treatment plant: MECs vs PECs?
Environ Int 2009;35(5):787–92.
De Young DJ, Bantle JA, Hull MA, Burks SL. Differences in sensitivity to developmental
toxicants as seen in Xenopus and Pimephales promelas embryos. Bull Environ Con-
tam Toxicol 1996;56(1):143–50.
EMEA. Note for guidance on environmental risk assessment of medicinal products for human
use. Doc. Ref. EMEA/CHMP/SWP/4447/00. Doc. Ref. EMEA/CHMP/SWP/4447/00. London,
UK: European Agency for the Evaluation of Medicinal Products; 2006.
Emmanuel E, Pierre MG, Perrodin Y. Groundwater contamination by microbiological
and chemical substances released from hospital wastewater: health risk assess-
ment for drinking water consumers. Environ Int 2009;35:718–26.
Ericson JF. An evaluation of the OECD 308 water/sediment systems for investigating the
biodegradation of pharmaceuticals. Environ Sci Technol 2007;41:5803–11.
Fabre G, Goldman ID. Formation of 7-hydroxymethotrexate polyglutamyl derivatives
and their cytotoxicity in human chronic myelogenous leukemia cells in vitro. Can-
cer Res 1985;45:80–5.
FASS. The Swedish medicines information engine. Läkemedelsfakta: Miljöinformation.
Last accessed march 2011. Available at www.fass.se.
FDA-CDER. Retrospective review of ecotoxicity data submitted in environmental as-
sessments. Rockville, MD, USA: FDA Center for Drug Evaluation and Research;
1996 (docket N 96N-0057).
Fent K, Weston AA, Caminada D. Ecotoxicology of human pharmaceuticals. Aquat Tox-
icol 2006;76(2):122–59.
Ferk F, Mišík M, Grummt T, Majer B, Fuerhacker M, Buchmann C, et al. Genotoxic effects
of wastewater from an oncological ward. Mutat Res 2009;672:69–75.
Frenzilli G, Nigro M, Lyons BP. The Comet assay for the evaluation of genotoxic impact
in aquatic environments. Mutat Res — RevMutat Res 2009;681(1):80–92.
Gust M, Garric J, Giamberini L, Mons R, Abbaci K, Garnier F, et al. Sensitivity of New
Zealand mudsnail Potamopyrgus antipodarum (Gray) to a specific aromatase in-
hibitor. Chemosphere 2010;79(1):47–53.
Halling-Sørensen B, Nors Nielsen S, Lanzky PF, Ingerslev F, Holten-Lützhøft HC,
Jørgensen SE. Occurrence, fate and effects of pharmaceutical substances in the en-
vironment — a review. Chemosphere 1998;36(2):357–93.
Henschel KP, Wenzel A, Diedrich M, Fliedner A. Environmental hazard assessment of
pharmaceuticals. J Reg Toxicol Pharmacol. 1997;25(3):220–5.
Hignite C, Aznaroff DL. Drugs and drugs metabolites as environmental contaminants:
chlorophenoxyisobutirate and salicylic acid in sewage effluent. Life Sci 1977;20:
337–41.
Jensen KM, Kahl MD, Makynen EA, Korte JJ, Leino RL, Butterworth BC, et al. Character-
ization of responses to the antiandrogen flutamide in a short-term reproduction
assay with the fathead minnow. Aquat Toxicol 2004;70:99-110.
Johnson AC, Jürgens MD, Williams RJ, Kümmerer K, Kortenkamp A, Sumpter JP. Do cytotoxic
chemotherapy drugs discharged into rivers pose a risk to the environment and human
health? An overview and UK case study. J Hydrol 2008;348(1–2):167–75.
Jolibois B, Guerbet M. Evaluation of industrial, hospital and domestic wastewater gen-
otoxicity with the Salmonella fluctuation test and the SOS chromotest. Mutat Res
2005;565(2):151–62.
Joqueviel C, Martino L, Gilard V, Malet Martino M, Canal P, Niemeyer U. Urinary excre-
tion of cyclophosphamide in humans determined by phosphorus-31 nuclear mag-
netic resonance spectroscopy. Drug Metab Dispos 1998;26:418–28.
Kiffmeyer T, Götze H-J, Jursch M, Lüders U. Trace enrichment, chromatographic separa-
tion and biodegradation of cytostatic compounds in surface water. Fresenius J Anal
Chem 1998;361(2):185–91.
Kisanga ER, Mellgren G, Lien EA. Excretion of hydoxylated metabolites of Tamoxifen in
human bile and urine. Anticancer Res 2005;25:4487–92.
Kolpin DW, Furlong ET, Meyer MT, Thurman EM, Zaugg SD, Barber LB, et al. Pharma-
ceuticals, hormones, and other organic wastewater contaminants in U.S. streams,
1999–2000: a national reconnaissance. Environ Sci Technol 2002;36:1202–11.
Kümmerer K. Pharmaceuticals in the environment. In: Kümmerer K, editor. Pharma-
ceuticals in the environment : sources, fate, effects and risks. Berlin: Springer-Ver-
lag; 2001. pp. 1–8.
Kümmerer K, Al-Ahmad A. Biodegradability of the anti-tumour agents 5-fluorouracil,
cytarabine and gemcitabine: impact of the chemical structure and synergistic tox-
icity with hospital effluent. Acta Hydroch Hydrob 1997;25:166–72.
Kümmerer K, Steger-Hartmann T, Meyer M. Biodegradability of the anti-tumour agent
ifosfamide and its occurrence in hospital effluents and communal sewage. Water
Res 1997;31(11):2705–10.
Kümmerer K, Al-Ahmad A, Bertram B, Wießler M. Biodegradability of antineoplastic
compounds in screening tests: influence of glucosidation and of stereochemistry.
Chemosphere 2000;40(7):767–73.
Kümmerer K, Alexy R, Hüttig J, Schöll A. Standardized tests fail to assess the effects of
antibiotics on environmental bacteria. Water Res 2004;38(8):2111–6.
Lacaze E, Geffard O, Bony S, Devaux A. Genotoxicity assessment in the amphipod
Gammarus fossarum by use of the alkaline Comet assay. Mutat Res 2010;700
(1–2):32–8.
85
Lacaze E, Geffard O, Goyet D, Bony S, Devaux A. Linking genotoxic responses in Gam-
marus fossarum germ cells with reproduction impairment, using the Comet
assay. Environ Res 2011;111(5):626–34.
Langford K, Thomas KV. Input of selected human pharmaceutical metabolites into the
Norwegian aquatic environment. J Environ Monit 2010;13(2):416–21.
Lenz K, Hann S, Koellensperger G, Stefanka Z, Stingeder G, Weissenbacher N, et al.
Presence of cancerostatic platinum compounds in hospital wastewater and pos-
sible elimination by adsorption to activated sludge. Sci Totat Environ 2005;345:
141–52.
Lenz K, Mahnik SN, Weissenbacher N, Mader RM, Krenn P, Hann S, et al. Monitoring,
removal and risk assessment of cytostatic drugs in hospital wastewater. Water
Sci Technol 2007;56:141–9.
Lingg RM, Hempel G, Rots MG, Van Zantwijk CH, Boos J, Kaspers GJL. Effects and inter-
action of 7-hydroxy methotrexate and methotrexate in leukaemic cells ex vivo
measured by the thymidylate synthase inhibition assay. Cancer Chemother Phar-
macol 2005;56(3):322–7.
Liu X, Zhang J, Yin J, Duan H, Wu Y, Shao B. Analysis of hormone antagonists in clinical
and municipal wastewater by isotopic dilution liquid chromatography tandem
mass spectrometry. Anal Bioanal Chem 2010;396(8):2977–85.
Ludwig DL, Pereira DS, Zhu Z, Hicklin DJ, Bohlen P. Monoclonal antibody therapeutics
and apoptosis. Oncogene 2003;22:9097–106.
Mahnik SN, Rizovski B, Fuerhacker M, Mader RM. Determination of 5-fluorouracil in
hospital effluents. Anal Bioanal Chem 2004;380(1):31–5.
Mahnik SN, Lenz K, Weissenbacher N, Mader RM, Fuerhacker M. Fate of 5-fluorouracil,
doxorubicin, epirubicin, and daunorubicin in hospital wastewater and their elimi-
nation by activated sludge and treatment in a membrane-bio-reactor system. Che-
mosphere 2007;66(1):30–7.
Manning G, Plowman GD, Hunter T, Sudarsanam S. Evolution of protein kinase signal-
ing from yeast to man. Trends Biochem Sci 2002;27:514–20.
Micromedex Drugdex©. Thomson micromedex©. Healthcare series. http:www.micro-
medex.com/products/drugdex/. 2011. Last accessed august 2011.
Moldovan Z. Occurrences of pharmaceutical and personal care products as micropollu-
tants in rivers from Romania. Chemosphere 2006;64(11):1808–17.
Mullot J-U. Modélisation des flux de médicaments dans les effluents hospitaliers. 2009.
PhD thesis. University of Paris-Sud 11, France. Available at www.lspe.u-psud.fr/
These%20Ju%20Mullot.pdf.
Mullot J-U, Karolak S, Fontova A, Levi Y. Modeling of hospital wastewater pollution by
pharmaceuticals: first results of Mediflux study carried out in three French hospi-
tals. Water Sci Technol 2010;62(12):2912–9.
Panter GH, Hutchinson TH, Hurd KS, Sherren A, Stanley RD, Tyler CR. Successful detec-
tion of (anti-)androgenic and aromatase inhibitors in pre-spawning adult fathead
minnows (Pimephales promelas) using easily measured endpoints of sexual devel-
opment. Aquat Toxicol 2004;70(1):11–21.
Prado T, Silva DM, Guilayn WC, Rose TL, Gaspar AMC, Miagostovich MP. Quantifica-
tion and molecular characterization of enteric viruses detected in effluents
from two hospital wastewater treatment plants. Water Res 2011;45(3):
1287–97.
Preston BL, Snell TW, Robertson TL, Dingmann BJ. Use of freshwater rotifer Brachionus
calyciflorus in screening assay for potential endocrine disruptors. Environ Toxicol
Chem 2000;19(12):2923–8.
Reifferscheid G, Grummt T. Genotoxicity in German surface waters — results of a col-
laborative study. Water Air Soil Pollut 2000;123(1–4):67–79.
Roberts PH, Thomas KW. The occurrence of selected pharmaceuticals in wastewater ef-
fluent and surface waters of the lower Tyne catchment. Sci Totat Environ
2006;356:143–53.
Rowney NC, Johnson AC, Williams RJ. Cytotoxic drugs in drinking water: a prediction
and risk assessment exercise for the Thames catchment in the United Kingdom.
Environ Toxicol Chem 2009;28:2733–43.
Runnalls TJ, Margiotta-Casaluci L, Kugathas S, Sumpter JP. Pharmaceuticals in the
aquatic environment: steroids and anti-steroids as high priorities for research.
Hum Ecol Risk Assess 2010;16:1318–38.
Sárria MP, Santos MM, Reis-Henriques MA, Vieira NM, Monteiro NM. The unpredict-
able effects of mixtures of androgenic and estrogenic chemicals on fish early life.
Environ Int 2011;37(2):418–24.
Sebire M, Allen Y, Bersuder P, Katsiadaki I. The model anti-androgen flutamide sup-
presses the expression of typical male stickleback reproductive behaviour. Aquat
Toxicol 2008;90(1):37–47.
Seither RL, Rape TJ, Goldman ID. Further studies on the pharmacologic effects of the 7-
hydroxy catabolite of methotrexate in the L1210 murine leukemia cell. Biochem
Pharmacol 1989;38(5):815–22.
Steger-Hartmann T, Kümmerer K, Hartmann A. Trace analysis of the antineoplasics
ifosfamide and cyclophosphamide in sewage water by two step solid phase extrac-
tion and gas-chromatography–mass spectrometry. J Chromatogr A 1996;726
(1–2):179.
Straub JO. Combined environmental risk assessment for 5-fluorouracil and capecita-
bine in Europe. Integr Environ Assess Manag 2010;6:540–66.
Sun L, Zha J, Spear PA, Wang Z. Toxicity of the aromatase inhibitor letrozole to Japanese
medaka (Oryzias latipes) eggs, larvae and breeding adults. Comp Biochem Physiol C
Pharmacol Toxicol Endocrinol 2007;145(4):533–41.
Sun L, Zha J, Wang Z. Effects of binary mixtures of estrogen and antiestrogens on Japa-
nese medaka (Oryzias latipes). Aquat Toxicol 2009;93(1):83–9.
Ternes TA. Occurrence of drugs in German sewage treatment plants and rivers. Water
Res 1998;32:3245–60.
Thomas KV, Dye C, Schlabach M, Langford KH. Source to sink tracking of selected
human pharmaceuticals from two Oslo city hospitals and a wastewater treatment
works. J Environ Monit 2007;9(12):1410–8.
86 J.-P. Besse et al. / Environment International 39 (2012) 73–86
Trösken ER, Scholz K, Lutz RW, Volkel W, Zarn JA, Lutz WK. Comparative assessment of
the inhibition of recombinant human CYP19 (aromatase) by azoles used in agricul-
ture and as drugs for humans. Endocrinol Res 2004;30:387–94.
Trösken ER, Fischer K, Völkel W, Lutz WK. Inhibition of human CYP19 by azoles used as
antifungal agents and aromatase inhibitors, using a new LC–MS/MS method for the
analysis of estradiol product formation. Toxicology 2006:21933–40.
Urbatzka R, Watermann B, Lutz I, Kloas W. Exposure of Xenopus laevis tadpoles to finas-
teride, an inhibitor of 5-α reductase activity, impairs spermatogenesis and alters
hypophyseal feedback mechanisms. J Mol Endocrinol 2009;43(5):209–19.
Veltkamp SA, Pluim D, van Eijndhoven MAJ, Bolijn MJ, Ong FHG, Govindarajan R, et al.
New insights into the pharmacology and cytotoxicity of gemcitabine and 2′,2′-
difluorodeoxyuridine. Mol Cancer Ther 2008;7(8):2415–25.
Verlicchi P, Galletti A, Petrovic M, Barcelo D. Hospital effluents as a source of emerging
pollutants: an overview of micropollutants and sustainable treatment options.
J Hydrol 2010;389(3–4):416–28.
Williams TD, Caunter JE, Lillicrap AD, Hutchinson TH, Gillings ED, Duffell S. Evaluation
of the reproductive effects of tamoxifen citrate in partial and full life-cycle studies
using fathead minnows (Pimephales promelas). Environ Toxicol Chem 2007;26(4):
695–707.
Wu X, Hawse JR, Subramaniam M, Goetz MP, Ingle JN, Spelsberg TC. The tamoxifen me-
tabolite, endoxifen, is a potent antiestrogen that targets estrogen receptor a for
degradation in breast cancer cells. Cancer Res 2009;69(5):1722–7.
Yin J, Shao B, Zhang J, Li K. A preliminary study on the occurrence of cytostatic drugs
in hospital effluents in Beijing, China. Bull Environ Contam Toxicol 2010;84:
39–45.
Zheng Y, Sun D, Sharma AK, Chen G, Amin S, Lazarus P. Elimination of antiestrogenic
effects of active tamoxifen metabolites by glucuronidation. Drug Metab Dispos
2007;35:1942–8.
Zounkova R, Kovalova L, Blaha L, Dott W. Ecotoxicity and genotoxicity assessment of
cytotoxic antineoplastic drugs and their metabolites. Chemosphere 2010;81(2):
253–60.
Zounková R, Odráška P, Doležalová L, Hilscherová K, Maršálek B, Bláha L. Ecotoxicity
and genotoxicity assessment of cytostatic pharmaceuticals. Environ Toxicol Chem
2007;26(10):2208–14.
Zuccato E, Castiglioni S, Fanelli R. Identification of the pharmaceuticals for human use
contaminating the Italian aquatic environment. J Hazard Mater 2005;122(3):
205–9.