Environmental Pollution 233 (2018) 356e363

Contents lists available at ScienceDirect

Environmental Pollution
journal homepage: www.elsevier.com/locate/envpol

Evaluation of acute and chronic ecotoxicity of cyclophosphamide,

ifosfamide, their metabolites/transformation products and UV treated
samples*

Chiara Russo a, Margherita Lavorgna a, Marjeta Cesen b, c
, Tina Kosjek b, c,
b, c, ** a, *
Ester Heath , Marina Isidori
a  della Campania L. Vanvitelli, Via Vivaldi 43, I-81100 Caserta, Italy
Dipartimento di Scienze e Tecnologie Ambientali, Biologiche e Farmaceutiche, Universita
b
Jozef Stefan International Postgraduate School, Jamova 39, 1000 Ljubljana, Slovenia
c
Jozef Stefan Institute, Department of Environmental Sciences, Jamova 39, 1000 Ljubljana, Slovenia

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

Article history: Cyclophosphamide (CP) and Ifosfamide (IF) are two nitrogen mustard drugs widely prescribed in cancer
Received 28 July 2017 therapy. They are continuously released via excreta into hospital and urban wastewaters reaching
Received in revised form wastewater treatment plants. Although CP and IF, their metabolites and transformation products (TPs)
12 October 2017
residues have been found in the aquatic environment from few ng L
1 to tens of mg L
1, their environ-
Accepted 15 October 2017
Available online 5 November 2017
mental toxic effects are still not well known. The present study aimed to investigate the acute and
chronic ecotoxicity of CP and IF and their commercially available human metabolites/TPs, i.e. carboxy-CP,
Keto-CP and N-dechloroethyl-CP on different organisms of the aquatic trophic chain. The experiments
Keywords:
Cyclophosphamide
were performed using the green alga Pseudokirchneriella subcapitata, the rotifer Brachionus calyciflorus
Ifosfamide and the crustaceans Thamnocephalus platyurus and Ceriodaphnia dubia. Moreover, to assess the treatment
Acute toxicity conditions in regards to parent compound removal and formation of new TPs, CP and IF were UV-
Chronic toxicity irradiated for 6 h, 12 h, 24 h, 36 h and 48 h, followed by toxicity evaluation of treated samples by algae,
Metabolites/TPs rotifers and crustaceans. Between the parent compounds, IF resulted as more toxic drug under tested
UV-treatment conditions, exerting both acute and chronic effects especially on C. dubia (LC50:196.4 mg L
1,
EC50:15.84 mg L
1). Among the tested metabolites/TPs, only carboxy-CP inhibited the reproduction in
the rotifer. However, LOEC and NOEC values were calculated for CP and IF for all organisms. In addition,
despite a low degradation of CP (28%) and IF (36%) after 48 h UV-irradiation, statistically significant effect
differences (p < 0.05) from not-irradiated and irradiated samples were observed in both acute and
chronic assays, starting from 6 h UV-irradiation. Our results suggest that the toxic effects found in the
aquatic organisms may be attributable to interactions between the parent compounds and their me-
tabolites/TPs.
© 2017 Elsevier Ltd. All rights reserved.

1. Introduction
Alkylating agents have been used in cancer therapy since the
1940s. They are known to directly interact with DNA, what causes
*
This paper has been recommended for acceptance by Klaus Kummerer.
* Corresponding author. Dipartimento di Scienze e Tecnologie Ambientali, Bio-
 della Campania L. Vanvitelli, Via Vivaldi 43, I-
logiche e Farmaceutiche, Universita
81100 Caserta, Italy.
** Corresponding author. Jozef Stefan International Postgraduate School, Jamova
39, 1000 Ljubljana, Slovenia.
E-mail addresses: Ester.Heath@ijs.si (E. Heath), marina.isidori@unicampania.it
(M. Isidori).
crosslinking and strand breaking of DNA and inhibition of cell di-
vision with consequent cell death (Ralhan and Kaur, 2007). Ac-
cording to ATC classification, anticancer drugs are divided into the
following groups: L01AA Nitrogen mustard analogues, L01AB Alkyl
sulfonates, L01AC Ethylene imines, L01AD Nitrosoureas, L01AG
Epoxides and L01AX Other alkylating agents. In this broad spec-
trum, nitrogen mustards, represent the most widely prescribed
chemotherapeutics used for the treatment of lymphoma, leukemia
and several solid tumors, where they nonspecifically alkylate the N7
of the guanine bases by the aziridinium group in tumor tissue (Xie,
2012). However, cytotoxic effects were observed towards healthy
cells with adverse effects such as carcinogenicity, teratogenicity,

https://doi.org/10.1016/j.envpol.2017.10.066
0269-7491/© 2017 Elsevier Ltd. All rights reserved.
 C. Russo et al. / Environmental Pollution 233 (2018) 356e363
fetotoxicity both in patients and in health care workers in reference
to the first applied nitrogen mustards Cyclophoshamide (CP,
L01AA01) and Ifosfamide (IF, L01AA06) (Lekskulchai, 2016). They
are polar compounds, highly soluble in water, with low octanol-
water partition coefficient (log Kow of 0.6 and 0.9 for CP and IF,
respectively, Table S1) and low adsorption potential to sewage
sludge or sediments, which suggest they can be mainly found in the
water phase (Kosjek and Heath, 2011; Xie, 2012).
Based on pharmacokinetics, CP and IF are pro-drugs, which are
metabolised and excreted together with their metabolites from
human body via urine and/or faeces. About 62% of the initial CP
dosage is excreted via urine unaltered together with its metabolites
when administration lasts over 48 h (Eitel et al., 2000). The total
amount of IF and its metabolites excreted over a 24 h period via
urine is ~ 50% of the injected IF dose. The excretion % of e.g.
carboxy-CP, keto-CP and N-dechloroethyl-CP relative to the dose
applied is 11e23%, 0.4e0.6% and 2e3% (as CP metabolite) and 11%
(as IF metabolite), respectively (Gilard et al., 1993; Joqueviel et al.,
1998; Zhang et al., 2006). These studies report that the excretion
depend on age, gender, physical condition of patient, the admin-
istered dosage, the duration of the intravenous administration and
the accompanying therapy with other pharmaceuticals.
As these residues mainly end up in sewage system, their main
contamination source is represented by wastewater treatment
plant effluents (Kosjek and Heath, 2011; Isidori et al., 2016).
The low removal efficiencies for both drugs in wastewater
treatment plants have raised concern for environmental health and
safety implications despite existence of various removal techniques
reported in literature such as biotransformation, ozonation, air
stripping, sorption and phototransformation. In fact, Buerge et al.
(2006) found that no degradation was observed in activated
sludge incubation experiments within 24 h at concentrations of
~100 ng L
1. Lutterbeck et al. (2016) also found that no degradation
was observed for CP using UV and simulated sunlight treatments.

Contrarily, Cesen et al. (2015) showed higher removal efficiencies
for these two compounds (59% and 35% for CP and IF, respectively)
when Moving Bed Biofilm Reactor (MBBR) technique was applied
(attached growth biomass) and even higher removals (99% for CP
and 94% for IF) when UV treatment and ozonation with addition of
H2O2 were sequentially included after biological treatment. In 2015,
Lin and colleagues investigated the removal efficiency of ozonation
for different anticancer drugs including CP and IF and reported the
resistance of process by-products to further ozonation. This resis-
tance could justify the increase in Microtox acute toxicity results
after only 30 min of ozonation despite the rapid degradation of CP
and IF which was caused by the attack of hydroxyl radicals.
Accordingly, CP and IF residues including residues of parent com-
pounds, metabolites and TPs have been found in wastewaters at
concentrations from pg L
1 to tens of mg L
1 (Buerge et al., 2006;

Cesen et al., 2015; Isidori et al., 2016).
Nonetheless, toxic effects of CP and IF metabolites/TPs are not

well understood. Cesen et al. (2016a) identified four CP and four IF
TPs formed upon UV treatment and five CP and four IF TPs during
UV/H2O2 treatment. In addition, the authors measured CP and IF
selected metabolites/TPs, i.e. as carboxy-CP, keto-CP and N-dech-
loroethyl-CP (or structural analogue 3-dechloroethyl-IF) in hospital
wastewaters at concentrations from ng to tens of mg L
1. However,
to the author's knowledge, there is no evident toxic effects of both
parent compounds and metabolites/TPs reported in concentrations
below mg L
1 (Lin et al., 2015). Data on the long-term toxic effects
of CP, IF and their metabolites/TPs in non-target organisms are also
lacking.
The aim of this study was to evaluate the ecotoxicity of the
parent drugs CP and IF and their commercially available metabo-
lites/TPs: carboxy-CP, keto-CP and N-dechloroethyl-CP and its
357
structural analogue 3- dechloroethyl-IF. This paper follows the

research conducted by Cesen and collaborators (2016b) where the
toxic effects of these compounds have been evaluated on pro-
ducers, the alga Pseudokirchneriella subcapitata and the cyanobac-
teria Synecococcus leopoliensis. Since the evaluation of the possible
ecotoxicological effects on organisms from different trophic levels
is necessary for the environmental risk assessment of such com-
pounds, in the present study, the acute and the chronic toxicity of
CP, IF and the metabolites/TPs was tested in primary consumers.
The following organisms were utilized: the environmentally
abundant rotifer Brachionus calyciflorus, the anostracan crustacean
Thamnocephalus platyurus, which was used because of its high
sensitivity in acute toxicity testing and the cladoceran crustacean
Ceriodaphnia dubia, which is distributed in freshwaters worldwide
and can be employed in both, acute and chronic tests. Moreover,
growth inhibition of CP, IF and their metabolites/TPs was also
evaluated in the green alga Pseudokirchneriella subcapitata to obtain
their unknown LOEC and NOEC. Finally, to evaluate the treatment
efficiency of CP and IF, the effects of UV irradiated CP and IF was
evaluated by alga, rotifer and cladoceran crustacean toxicity testing
to assess the risk associated with unknown TP formation.
2. Materials and methods
2.1. Standards, reagents and chemicals
CP (99%, CAS 50-18e0 P) and IF (99%, CAS 3778-73-2) were
purchased from Sigma Aldrich (Hong Kong, China). Deuterated
cyclophosphamide (d6-CP), used as internal standard for chemical
analysis, carboxy-cyclophosphamide (CAS 22788-18-7), keto-CP
(CAS 27046-19-1) and N-decl-CP (a structural analogue of CP and
IF metabolites/TPs; CAS 36761-83-8) were obtained from Niomech
- IIT GmbH (Bielefeld, Germany). Derivatizing agent trifluoroacetic
anhydride (TFAA, 99%, CAS 407-25-0) was purchased from Fluka
(Buchs, Switzerland). Acetonitrile and ethyl acetate were purchased
from J. T. Baker (Deventer, Netherlands) and were of analytical
grade purity. To prevent scavenging during UV irradiation experi-
ments, stock solutions of CP and IF were prepared in MilliQ water.
Calibration standards were prepared by dilution of stock solutions
in MilliQ water.
2.2. UV irradiation experiment
Duplicate UV experiments were performed in a cylindrical glass
reactor containing 760 mL of aqueous solution with initial con-
centration of CP and IF at 200 mg L
1 in MilliQ water. The drugs
selected were tested at high concentrations, certainly far from
environmental concern, considering CP and IF occurrence in surface
waters, but necessary to obtain ecotoxicological data useful for
further risk assessment. The high concentrations were also selected

based on our previous study (Cesen et al., 2016b) and the study of
Sanderson et al. (2003) addressing the toxicity of CP and IF. Tested
solutions were continuously stirred by magnetic stirrer (200 rpm)
throughout each experiment. Monochromatic low pressure (LP)
mercury lamp was introduced into the reactor and separated from
testing solution by quartz cooling well, which maintained the so-
lution temperature at approximately 20  C. The LP lamp (6 W) as
irradiation source emits 90% of its radiation at 254 nm. The UV
intensity, which was measured by ferrioxalate actinometry, was
0.78  10
6 Einstein/sec (Nicole et al., 1990). Fresh MilliQ water was
used as medium because it has only negligible amounts of organic
matter (<5 ppb) so that the formation of various reactive oxygen
species (ROS) induced by UV irradiation could be excluded. The
MilliQ water with resistivity of 18.2 MU cm at 25  C which ensures
pH of 7.0 was applied. The pH of the media was not measured
358 C. Russo et al. / Environmental Pollution 233 (2018) 356e363
within the experiments in order to avoid microbial contamination.
In addition, to avoid any significant changes in the pH of the media
during UV experiments, experiments were performed in a closed
atmosphere to prevent the drop of pH, which could derive from
dissolved ambient CO2. The photodegradation efficiency of each
analyte was determined by the ratio between concentrations
determined in treated samples and in non-treated controls. One
hundred microliters and 20 mL of samples were collected after 6 h,
12 h, 24 h, 36 h and 48 h of UV irradiation for chemical analysis and
bioassays, respectively.
2.3. Sample preparation for chemical analysis
As the initial concentration within the UV experiments was
high, i.e. 200 mg L
1, preconcentration of samples was not needed.
Instead, 100 mL of samples including controls and UV-irradiated
samples were dried under nitrogen at 40  C. The dried samples
were derivatized for 1 h at 60  C with 100 mL of TFAA and 100 mL of
ethyl acetate. The samples were then dried again under nitrogen to
remove any trifluoroacetic acid formed during derivatization and
re-dissolved in 250 mL of ethyl acetate prior to chemical analysis.
2.4. Chemical analysis
The samples were analysed using a HP 6890 series (Hewlett-
Packard, Waldbron, Germany) gas chromatograph with a single
quadrupole mass selective detector. The capillary column was a DB-
5MS 30 m  0.25 mm  0.25 mm (Agilent J&W, CA, US). The carrier
gas was He (flow rate of 1 mL/min). One microlitre of sample
extract was injected in splitless mode with the injector tempera-
ture set at 270  C. The temperature program was as follows: an
initial temperature of 65  C was ramped at 30  C/min to 300  C,
where it was held for 2 min. Total runtime was 9.83 min. The MS
was operated in electron impact (EI) mode with an ionization
voltage of 70 eV. Selected ion monitoring (SIM) mode was used for
qualitative and quantitative analysis by monitoring the following
ions: m/z 309, 307 and 150 for CP, m/z 307, 181 and 150 for IF and
m/z 313 and 315 for CP-d6. Instrumental control and data pro-
cessing were performed using ChemStation software.
2.5. Toxicity tests
2.5.1. Acute toxicity tests
B. calyciflorus acute assay was performed following the ASTM E
1440e91 guidelines. Organisms less than 2 h old were hatched
from cysts (MicroBioTest Inc., Nazareth, Belgium) in synthetic
moderately hard freshwater (80e100 mg L
1 CaCO3, pH 7.5 ± 0.3) at
25 ± 1  C under continuous illumination (3000e4000 lux) for
16e18 h prior to test initiation. 0.3 mL of test solution was used in
each test well (36-well plates, MicroBioTest Inc., Nazareth,
Belgium). Single compounds (in four to five dilutions starting from
the highest concentrations of 2000 mg L
1 and 100 mg L
1 for
parent compounds and metabolites, respectively, with a geometric
progression of 2) as well as UV-irradiated samples at different
exposure times (6 h, 12 h, 24 h, 36 h and 48 h) were tested in six
replicates with five animals/well.
T. platyurus test (ISO 14380, 2011) was conducted on larvae
hatched from cysts 20e22 h before the assay in the same synthetic
freshwater as the rotifers and under the same experimental con-
ditions. Tests were performed in 24-well plates using 10 crusta-
ceans/well, 1.0 mL in five dilutions starting from the highest
concentrations of 2000 mg L
1 and 100 mg L
1 for parent com-
pounds and metabolites respectively (geometric progression 2) and
in three replicates.
C. dubia test was performed following EPA-600-4-90 procedures
(US EPA, 1993) using less than 24 h old organisms. Neonates of at
least third generation came from the mass culture (Aquatic
Research Organisms, Inc., Hampton, NH, USA) and were maintained
at 25 ± 1  C in synthetic ISO medium (hardness 250 mg L
1
expressed as CaCO3) with a 16:8 h light:dark cycle (600 lux). Tests
were performed in 24-well plates with 10 crustaceans/well and
with either 1.0 mL of the four to seven dilutions of single com-
pounds (starting from the highest concentrations of 2000 mg L
1
and 100 mg L
1 for parent compounds and metabolites respec-
tively, geometric progression 2) or with UV-irradiated samples at
different exposure times. The tests were performed in three
replicates.
For all tests, a test-medium control series (negative control) was
used to the test series. Subsequently, all plates were incubated in
darkness at 25  C for 24 h for each test.
Mortality was the end-point considered in the acute tests and
the concentration resulting in a 50% effect after 24 h-exposure was
indicated as Median Lethal Concentration (LC50).
2.5.2. Chronic toxicity tests
B. calyciflorus chronic test was based on the offspring reduction
over 48 h exposure (ISO, 20666, 2008) and was performed on less
than 2 h old organisms. Cysts were hatched as described for the
acute test. Tests were performed in 48-well plates with one rotifer/
well and with either 0.9 mL/well of single compounds (in six di-
lutions in moderately hard dilution water, starting from the highest
concentrations of 200 mg L
1 and 100 mg L
1 for parent com-
pounds and metabolites, respectively, with a geometric progression
of 2) or with different UV-irradiated samples. All tests were per-
formed in six replicates. The organisms were fed with 0.1 mL of
fresh suspension of 107 cells/mL of the unicellular alga
P. subcapitata. Plates were incubated in darkness at 25  C for 48 h.
The test on C. dubia 24 h old neonates was performed over 7
days (ISO, 20665, 2008). Organisms from at least the third gener-
ation were individually exposed to 25 mL of test solution in bea-
kers. All test media were exchanged five times per week in semi-
static conditions. From the fourth day-exposure onward, the
offspring produced by each parent animal were counted and
removed daily. Crustaceans were fed with a combination of yeast
Saccharomyces cerevisiae, alfalfa and flake food (5 g L
1 of each) in
addition to the unicellular green alga P. subcapitata (108 cells/mL).
Single compounds (in five to seven dilutions starting from the
highest concentrations of 200 mg L
1 and 10 mg L
1 for parent
compounds and metabolites/TPs, respectively, with a geometric
progression of 2) as well as UV-irradiated samples were tested in
ten replicates and incubated at 25 ± 1  C, with a 16:8 h light:dark
cycle (600 lux).
P. subcapitata growth inhibition test (Paixao et al., 2008) was
performed in 96-well microplates. The single compounds (six-
seven dilutions starting from the highest concentrations of
200 mg L
1 and 10 mg L
1 for parent compounds and metabolites/
TPs, respectively, with a geometric progression 2) and the UV-
irradiated samples were incubated with 104 cells/mL of algal sus-
pension in six replicates under continuous illumination (6000 lux)
at 25 ± 1  C on a microplate shaker (450 rpm). The plates were read
at 450 nm (SpectraFluor, Tecan, Switzerland) immediately before
the test and after 24 h, 48 h and 72 h.
For all chronic tests, a test-medium control series (negative
control) was used to the test series. The number of the offspring or
the algal growth outputs were compared to the values obtained for
the negative control in order to determine the chronic effective
percentages and to evaluate the chronic Effective Concentrations
(ECx).
 C. Russo et al. / Environmental Pollution 233 (2018) 356e363
2.5.3. Data analysis
For each test, two (for UV-irradiated samples) or three (for
parent compounds and metabolites/TPs) independent experiments
were performed.
C. dubia independent assays were analysed using ToxRat Pro-
fessional software, Ver 2.10.05 (Alsdorf, Germany) for the calcula-
tion of effective percentages.
For all organisms the effective percentages obtained from in-
dependent experiments were pooled using Prism5 software
(Graphpad Inc., CA, USA) to estimate the concentrations giving x%
effect (L(E)Cx) by non-linear regression (log agonist vs. normalized
response-variable slope). The LC50 value, corresponding to 50%
mortality, was the test parameter for organisms in the acute tests,
whereas EC50, EC20 and EC10 were the concentrations giving 50%,
20% or 10% of the effect used in chronic tests to evaluate the inhi-
bition reproduction effect or the inhibition of the algal growth.
For long-term toxicity, the Lowest Observed Effect Concentra-
tion (LOEC) and the No Observed Effect Concentration (NOEC) were
estimated by ANOVA and Dunnett's multiple comparison test.
Thus, for the inhibition of the reproduction tests, the compari-
son was evaluated between the mean number of live offspring
produced per parent organism, which was exposed to chemicals
and the offspring mean of the controls. Regarding the algal growth
inhibition tests, the comparison was between the mean number of
the absorbance values related to different compounds (directly
proportional to the number of algal cells) and the control absor-
bance mean values (after 72 h).
As regards to UV-irradiated samples, the statistically significant
differences between the non-irradiated and irradiated samples (at
different UV exposure times) as well as the significant difference
between different UV exposure times and between different or-
ganism species were evaluated by 2-WAY ANOVA, Bonferroni
posttests.
3. Results and discussion
3.1. Results of chemical analysis
Chemical analysis of UV-treated samples revealed poor degra-
dation of both investigated compounds, what was expected ac-
cording to their UV-Vis spectra in the range of 250e370 nm (Lin
et al., 2014). After 48 h, CP and IF were degraded only up to 28.3%
and 36.5%, respectively, (Table S2, Fig. 1). Degradation of both
compounds followed pseudo first order kinetics using the following
Equation (1):
Fig. 1. Degradation of CP and IF during UV-irradiation of aqueous samples (n ¼ 2).
359
cðtÞ
ln ¼
kt  t (1)
cð0Þ
where kt is the first-order rate constant including both direct and
indirect oxidation of the investigated compounds (Rosenfeldt and
Linden, 2007). The R2 values of degradation curves were 0.983
and 0.963 for CP and IF, respectively. The kt values for CP and IF
were 7.00 ± 1.43  10
3 h
1 (or 1.97  10
6 sec
1) and
1.04 ± 0.202  10
2 h
1 (or 2.9  10
6 sec
1), respectively. In
addition, poor removal was expected since chemical structures of
CP and IF do not contain double bonds that could absorb photons

under UV irradiation (Cesen et al., 2015). The kt value for CP
determined within this study is of similar order of magnitude as
previously reported by Kim and Tanaka (2009) in ultrapure water
(kt for CP ¼ 8.3  10
5 sec
1). Despite the similarity in obtained kt
values, the potential formation of radical species during photolysis
most probably contributed towards the observed CP and IF removal.
Another very important parameter for photodegradation to be
considered is the quantum yield of CP and IF in the UV 254 nm, as is
molar extinction coefficient (ε). A recent study by Zhang et al.
(2017) reports quantum yields for CP and 5-fluorouracil (both cy-
tostatics), which are 54.83  103 mol E
1 and 4.06  103 mol E
1,
respectively. The authors correlate low removal during sole UV
irradiation using LP lamp (5 W, UV irradiation at 254 nm) to low ε
for CP (7 M
1 cm
1), if compared to 5-fluorouracil
(3506 M
1 cm
1), which was degraded approximately 35-times
faster. They relate higher removal by high absorbance of 5-
fluorouracil at UV-254 nm and not to lower quantum yield. Based
on our results, we assume that the observed removal of CP and
structurally very similar IF might be attributed to low ε and to
potential formation of radical species formed upon treatment.
Furthermore, also the mineralization during photolysis is negli-
gible, what was confirmed in our previous study (Cesen  et al.,
2016b). In there, DOC was monitored in UV irradiated samples
containing CP and IF (at 10 mg L
1) and revealed no significant
decrease in DOC over 48 h of UV irradiation. This suggested that UV
degradation failed to mineralize both compounds.
3.2. Results of toxicological tests
3.2.1. Acute and chronic toxicity of parent compounds and their
metabolites/TPs
Table 1 shows acute toxicity of CP and IF expressed as LC50. For
both drugs, the median mortality of all invertebrates was found at
concentrations orders of magnitude higher (196.4e1924 mg L
1)
360 C. Russo et al. / Environmental Pollution 233 (2018) 356e363

Table 1 concentrations are too far from those of environmental concern

LC50 values in mg L
1 for acute toxicity tests with 95% confidence range.
Compounds C. dubia 24 h B. calyciflorus 24 h T. platyurus 24 h
CP 986.6 1924 1396
(765.3e1272) (1210e3060) (1304e1494)
IF 196.4 996.3 771.5
(149.1e258.7) (767.3e1294) (627.9e947.8)
For metabolites/TPs no effect was observed up to concentration of 100 mg L
1.
than concentrations found in the aquatic environment (Cesen  et al.,
2015; Isidori et al., 2016). Between the two parent drugs, IF was
more toxic for all non-target organisms, whose susceptibility was
found in the following order: B. calyciflorus < T. platyurus < C. dubia
highlighting that IF acute toxic effects increased not only as the
aquatic trophic chain level increased, but also with the increased
stages of phylogenetic tree. This is in accordance with a study by
Sanderson et al. (2003), where effects of the same antineoplastic
drugs were evaluated towards algae, daphnids and fish. The highest
toxic effects were reported for fish Pimephales promelas, i.e. the
organism with the highest evolutionary stage among the tested
organisms. The absence of acute toxicity for CP is also confirmed by
Lutterbeck et al. (2014), who did not find any luminescence inhi-
bition after 30 min in Vibrio fischeri. In the present study, CP and IF
metabolites/TPs did not reveal acute toxic effects up to 100 mg L
1,
which was the highest tested concentration.
In chronic toxicity tests (Table 2, Fig. 2), IF was confirmed to be
the most toxic compound causing a median offspring reduction in
C. dubia at 15.84 mg L
1 after 7d-exposure and in B. calyciflorus at
76.05 mg L
1 after 48 h-exposure. CP EC50 was 58.03 mg L
1 in C.
dubia with an EC10 of 19.50 mg L
1. This result is partially in
agreement with the EC10 found by Lutterbeck et al. (2014). An
increasing growth inhibition effect was observed for algae at con-
centrations up to the highest concentration tested of 200 mg L
1 for
both parent compounds. These findings are in agreement with a
study by Zounkova et al. (2007), who found a median growth in-
hibition in P. subcapitata at much higher concentrations
(930 mg L
1) for CP. An evident increase of the effect caused by IF
was observed both in acute and chronic toxicity moving up the
trophic chain. This suggests an increase of its toxic effect with the
increased complexity of organisms. According to Zhu et al. (2015),
different metabolic pathways could be linked to alkylating agents
metabolism such as GSH metabolism, fatty acid metabolism,
biosynthesis of unsaturated fatty acids, fatty acid biosynthesis and
steroid hormone biosynthesis pathways that could be involved in
bioaccumulation and bioconcentration processes also of human
concern.
Although for P. subcapitata the median toxic effect was not
experimentally reached, both, LOEC and NOEC values, were deter-
mined in the mg L
1 range (Table S3). Even though these observed
(mg-ng L
1) and are dependent on the dilution factor chosen in the
experimental plan, from a toxicological point of view the study of
the lowest effective and non-effective levels is a very useful tool for
further studies regarding the environmental risk assessment and
the calculation of the Risk Quotient (R.Q.) of the compounds here
studied. Furthermore, as observed only in case of CP (Fig. 2), both,
P. subcapitata and C. dubia, increased their growth/reproduction
rate at the lowest tested concentrations confirming adaptation
potential known as hormesis. This useful biological response to
toxins and stressors generally allows organisms to react to toxic
compounds. The hormetic doseeresponse curves, observed from
the cellular level to broad categories of organisms, compensate any
imbalance in homeostasis caused by mild stresses. For example,
reactive oxygen species are one of the major reasons for most of the
diseases, but exposure to low quantities reverses its negative effects
(Ristow and Schmeisser, 2014). Interestingly, in 2015 Gaya et al.
observed that the administration of CP at high doses exerted anti-
tumoral effect with bone marrow suppression as an adverse ef-
fect while, at low doses, it contributed to anti-tumor immunity.
Despite similarity in chemical structure, only carboxy-CP
showed chronic effects among the tested metabolites/TPs. In fact,
carboxy-CP caused inhibition of the reproduction in rotifers in a
concentration dependent manner (Fig. 3) with a median offspring
reduction equal to 23.66 mg L
1 (Table 2). The toxicity of carboxy-
CP occurring in the range of mg L
1 has already been reported by

Cesen et al. (2016b), who observed a median toxic effect at
17.1 mg L
1 with the cyanobacterium Synecococcus leopoliensis. In
line with the EU Directive 93/67/EEC (1996), which classifies sub-
stances according to their EC50 values towards aquatic organisms,
those authors defined carboxy-CP as harmful to aquatic organisms
having an EC50 in the range from 10 to 100 mg L
1. Since both
parent compounds showed an EC50 in the range from 10 to
100 mg L
1, when tested in C. dubia and B. calyciflorus (Table 2), also
CP and IF can be considered harmful to aquatic organisms. Besides,
these alkylating agents are known as mechlorethamine com-
pounds, having a bischloroethyl group that reacts, through an
aziridinium intermediate, with electron-rich atoms in macromol-
ecules to form highly reacting covalent bonds that cause toxic ef-
fects (Siddick, 2002).
3.2.2. Acute and chronic toxicity of CP and IF after UV-irradiation
The risk associated with unknown TP formation after UV-
irradiation was assessed by the effect-driven approach. In this
approach, the irradiated mixture of CP and IF underwent toxicity
testing to investigate their ecotoxicity at different UV exposure
times.
Acute toxicity results of UV-irradiated samples containing CP
and IF are shown in Fig. 4, where the mortality percentage obtained

Table 2
EC50, EC20, EC10 values (mg L
1) for chronic toxicity tests with 95% confidence range.

Compounds C. dubia 7d B. calyciflorus 48 h

EC50 EC20 EC10 EC50 EC20 EC10

CP 58.03 28.85 19.50 89.84 14.19 4.82

(37.43e89.98) (15.60e48.08) (9.57e38.64) (67.62e119.4) (7.87e22.18) (0e10.19)
IF 15.84 5.58 3.03 76.05 14.93 5.75
(13.03e19.27) (3.78e7.79) (0e5.00) (48.04e120.4) (6.59e27.10) (0e14.93)
Carboxy-CP 23.66 3.99 1.41
(15.45e36.23) (0e7.57) (0e3.93)
N.D. up to 10
Keto-CP
N.E. up to 100
N-dechloroethyl-CP

N.D.: Not determinable.

N.E.: No effect.
 C. Russo et al. / Environmental Pollution 233 (2018) 356e363 361

Fig. 2. Concentration/effect curves (non-linear regression: log agonist vs. normalized response-variable slope) of CP and IF in C. dubia, B. calyciflorus and P. subcapitata. Bars indicate
standard errors.

Fig. 3. Concentration/effect curves (non-linear regression: log agonist vs. normalized
response-variable slope) of Carboxy-CP, Keto-CP and N-dechloroethyl-CP in
B. calyciflorus. Bars indicate standard errors.
in B. calyciflorus and in C. dubia after 24 h exposure is given. In
addition, at each study time (6 h, 12 h, 24 h, 36 h and 48 h), a
statistically significant difference among different samples was
calculated. CP UV-irradiated sample showed the lowest toxicity
after irradiation in C. dubia reaching a significant effect (p < 0.05,
compared to non-irradiated samples) equal to 23.3% of mortality
after 48 h of UV-irradiation. CP with B. calyciflorus as well as IF with
B. calyciflorus and with C. dubia caused significant mortality rates
(compared to non-irradiated samples) of 26.7% (p < 0.01), 41.7%
(p < 0.001) and 74.8% (p < 0.001), respectively, after only 6 h UV-
treatment, reaching 95%, 96.7% and 98% of mortality after 48 h of
UV-irradiation. The significant mortality increase following the
exposure time is shown in Fig. 5. A clear time exposure-mortality
trend was observed in the rotifer exposed to irradiated CP,
differing from the concentration effect found when exposing the
crustacean to irradiated samples containing CP with no significant
increasing mortality. IF irradiated samples showed the highest
acute toxicity to both exposed organisms. Probably, the acute toxic
potential could be due to a mixture of different TPs formed during
treatment such as the human metabolites and TPs IF-TP199 (N-
dechloroethyl-ifosfamide), IF-TP275 (keto-ifosfamide), IF-TP138
and IF-TP259 (imino-ifosfamide) already tested by Cesen  et al.
(2016a) on Synecococcus leopoliensis.
Chronic toxicity results are shown in Fig. 6, where the offspring
reduction and the growth inhibition percentages observed for
B. calyciflorus/C. dubia and for P. subcapitata, respectively, are
correlated to the CP and IF UV-exposure time. The least toxic effect
was observed in B. calyciflorus exposed to UV-irradiated sample
containing CP, with initial significant inhibition of the reproduction
(81.05%) obtained after 24 h of UV-irradiation. The highest initial
toxicity level (p < 0.01) was found in B. calyciflorus exposed to 6 h-
UV irradiated sample containing IF (toxic effect of 86.35%). Chronic
effects in C. dubia of UV-treated samples containing IF are shown in
Fig. 7. The highest toxic result (77.95% offspring reduction) was
obtained with 1/10 dilution of IF irradiated by UV for 48 h, while an
initial significant toxic effect (p < 0.05) was found when testing 1/
1000 dilution after 6 h of UV irradiation. Figure S1 shows the

Fig. 4. Mortality percentage obtained testing UV-irradiated samples in B. calyciflorus and C. dubia. Significant differences were noted on comparing at each time non-irradiated and
irradiated samples for *p < 0.05; **p < 0.01; ***p < 0.001(2WAY ANOVA, Bonferroni posttests). Different letters were used for each significantly different sample (p < 0.05).
362 C. Russo et al. / Environmental Pollution 233 (2018) 356e363

Fig. 5. Mortality percentage/UV-irradiation time trend obtained in B. calyciflorus and C. dubia. Different letters for p < 0.05(2WAY ANOVA, Bonferroni posttests). Different numbers
were used to identify different samples.

Fig. 6. Offspring reduction and growth inhibition percentages observed respectively for B. calyciflorus/C. dubia and for P. subcapitata exposed to different times of samples UV-
irradiated. Significant differences were noted on comparing at each time non-irradiated and irradiated sample for *p < 0.05; **p < 0.01; ***p < 0.001(2WAY ANOVA, Bonferroni
posttests).

growth inhibition) is related to 36 h of UV exposure, while for CP

(85.16%) it is related to 24 h-exposure. The highest CP toxic effect in
rotifers and in crustaceans was related to 36 h of UV exposure
(96.30 and 88.85% inhibition of the reproduction, respectively),
whereas the first most significant toxic effect of IF was 94.55% in the
rotifer for a sample that was UV-irradiated for 24 h (Figure S1b).
The acute toxicity results were mostly related to the highest
toxic level in IF UV-irradiated samples, while the chronic outcomes
revealed also the ability of CP UV-irradiated samples to restrain
reproduction and exert growth inhibition. CP and IF were degraded
only up to 28% and 36%, respectively, but the findings of this study
suggest that a mixture of parent compounds and metabolites/TPs
could represent a harmful environmental combination exhibiting
synergistic and/or potentiation effects. Since the UV irradiated
mixtures showed an increase in toxicity at the same time as the
decrease in parent compound concentrations, further studies
where TPs will be separated, identified and quantified before
Fig. 7. Offspring reduction percentage testing IF UV-irradiated samples in C. dubia.
toxicity testing, as recommended by exposure-driven approach,
Different letters for p < 0.05(2WAY ANOVA, Bonferroni posttests). Different numbers
were used to identify different dilutions.
what is needed to confirm these assumptions.

chronic effect percentage/UV-irradiation time trend obtained in 4. Conclusions

both P. subcapitata (a) and in B. calyciflorus and C. dubia (b). In
Figure S1a, the highest toxic effect of IF in P. subcapitata (95.67% The findings of the present study reveal that the toxic effects of
 C. Russo et al. / Environmental Pollution 233 (2018) 356e363
the tested parent compounds occurred at concentrations several
orders of magnitude higher (mg L
1) than their environmental
relevant concentrations (ng-mg L
1). However, wastewater treat-
ment, metabolic and environmental transformations could lead to
the formation of a mixture of parent compound residues and me-
tabolites/TPs representing a harmful combination for the environ-
ment and then for human health. The UV-irradiation treatments,
here performed, highlighted the increase of toxicity due to the
presence of these TPs. In this study, the only metabolite/TP showing
a toxic potential was carboxy-CP although the synergies among the
other metabolites/TPs are still unknown and further studies are
needed to contribute to the environmental risk assessment.
Acknowledgements
This work was financially supported by the EU through the EU
FP7 project CytoThreat (Fate and effects of cytostatic pharmaceu-
ticals in the environment and the identification of biomarkers for
an improved risk assessment on environmental exposure. Grant
agreement no. 265264). Also, financial support from Slovene
Research Agency is kindly acknowledged (P1-0143, L1-5457, J1-
6752, L1-7544).
Appendix A. Supplementary data
Supplementary data related to this article can be found at
https://doi.org/10.1016/j.envpol.2017.10.066.
References
ASTM E 1440e1491, 2004. Standard Guide for Acute Toxicity with the Rotifer Bra-
chionus. American Society for Testing and Materials, Philadelphia PA, USA
reapproved.
Buerge, I.J., Buser, H.R., Poiger, T., Müller, M.D., 2006. Occurrence and fate of the
cytostatic drugs cyclophosphamide and Ifosfamide in wastewater and surface
waters. Environ. Sci. Technol. 40 (23), 7242e7250.

Cesen, M., Kosjek, T., Laimou-Geraniou, M., Kompare, B., Sirok,  B., Lambropolou, D.,
Heath, E., 2015. Occurrence of cyclophosphamide and ifosfamide in aqueous
environment and their removal by biological and abiotic wastewater treatment
processes. Sci. Total Environ. 15 (527e528), 465e473.

Cesen, M., Kosjek, T., Busetti, F., Kompare, B., Heath, E., 2016a. Human metabolites
and transformation products of cyclophosphamide and ifosfamide: analysis,
occurrence and formation during abiotic treatments. Environ. Sci. Pollut. Res. 23
(11), 11209e11223.

Cesen, 
M., Elersek, T., Novak, M., Zegura, B., Kosjek, T., Filipi
c, M., Heath, E., 2016b.
Ecotoxicity and genotoxicity of cyclophosphamide, ifosfamide, their metabo-
lites/transformation products and their mixtures. Environ. Pollut. 210, 192e201.
Eitel, A., Scherrer, M., Kümmerer, K., 2000. Handling Cytostatic Drugs: a Practical
Guide. Bristol-Myers-Squibb (2000).
Gaya, A., Akle, C.A., Mudan, S., Grange, J., 2015. The concept of hormesis in cancer
therapye is less more? Cureus 7 (4), 261.
Gilard, V., Malet-Martino, M.C., de Forni, M., Niemeyer, U., Ader, J.C., Martino, R.,
1993. Determination of the urinary excretion of ifosfamide and its phospho-
rated metabolites by phosphorus-31 nuclear magnetic resonance spectroscopy.
Cancer Chemother. Pharmacol. 31 (5), 387e394.

Isidori, M., Lavorgna, M., Russo, C., Kundi, M., Zegura, B., Novak, M., Filipi c, M.,
Misík, M., Knasmueller, S., Lopez de Alda, M., Barcelo 
 , D., Zonja, 
B., Cesen, M.,

S cancar, J., Kosjek, T., Heath, Ester, 2016. Chemical and toxicological character-
isation of anticancer drugs in hospital and municipal wastewaters from
Slovenia and Spain. Environ. Pollut. 219, 275e287.
363
ISO 14380, 2011. Water Quality-Determination of the Acute Toxicity to Thamnoce-
phalus Platyurus (Crustacea, Anostraca). International Organization for Stan-
dardization, Geneva, Switzerland.
ISO 20665, 2008. Water Quality-determination of Chronic Toxicity to Ceriodaphnia
Dubia in 7 Days-population Growth Inhibition Test. International Organization
for Standardization, Geneva, Switzerland.
ISO 20666, 2008. Water Quality-determination of Chronic Toxicity to Brachionus
Calyciflorus in 48 H-population Growth Inhibition Test. International Organi-
zation for Standardization, Geneva, Switzerland.
Joqueviel, C., Martino, R., Gilard, V., Malet-Martino, M., Canal, P., Niemeyer, U., 1998.
Urinary excretion of cyclophosphamide in humans, determined by Phosphorus-
31 nuclear magnetic resonance spectroscopy. Drug Metabolism Dispos. 26 (5),
418e428.
Kim, I., Tanaka, H., 2009. Photodegradation characteristics of PPCPs in water with
UV treatment. Environ. Int. 35 (5), 793e802.
Kosjek, T., Heath, E., 2011. Occurrence, fate and determination of cytostatic phar-
maceuticals in the environment. Trends Anal. Chem. 30 (7), 1065e1087.
Lekskulchai, V., 2016. Quantitation of anticancer drugs-Cyclophosphamide and
Ifosfamide in urine and water sewage samples by gas chromatography-mass
spectrometry. Int. J. Occup. Med. Environ. Health 29 (5), 815e822.
Lin, A.Y., Lin, Y.C., Lee, W.N., 2014. Prevalence and sunlight photolysis of controlled
and chemotherapeutic drugs in aqueous environments. Environ. Pollut. 187,
170e181.
Lin, A.Y., Hsueh, J.H., Hong, P.K., 2015. Removal of antineoplastic drugs cyclophos-
phamide, ifosfamide, and 5-fluorouracil and a vasodilator drug pentoxifylline
from wastewaters by ozonation. Environ. Sci. Pollut. Res. 22 (1), 508e515.
Lutterbeck, C.A., Kern, D.I., Machado, E.L., Kümmerer, K., 2014. Ecotoxicological
evaluation of selected anticancer drugs. Toxicol. Lett. 229, S71eS72.
Lutterbeck, C.A., Wilde, M.L., Baginska, E., Leder, C., Machado, E.L., Kümmerer, K.,
2016. Degradation of cyclophosphamide and 5-fluorouracil by UV and simu-
lated sunlight treatments: assessment of the enhancement of the biodegrad-
ability and toxicity. Environ. Pollut. 208, 467e476.
Nicole, I., De Laat, J., Dore, M., Duguet, J.P., Bonnel, C., 1990. Utilisation du rayon-
nementultraviolet dansle traitement des eaux: mesure duflux photonique par
actinometrie chimique au peroxyde d’hydrogene. Water Res. 24, 157e168.
Paixao, S.M., Silva, L., Fernandes, A., O'Rourke, K., Mendonca, E., Picado, A., 2008.
Performance of a miniaturized algal bioassay in phytotoxicity screening. Eco-
toxicology 17, 165e171.
Ralhan, R., Kaur, J., 2007. Alkylating agents and cancer therapy. Expert Opin. Ther.
Pat. 17 (9), 1061e1075.
Ristow, M., Schmeisser, K., 2014. Mitohormesis: promoting health and lifespan by
increased levels of reactive oxygen species (ROS). Dose-Response Publ. Int.
Hormesis Soc. 12, 288e341.
Rosenfeldt, E.J., Linden, K.G., 2007. The ROH,UV concept to characterize and the
model UV/H2O2 process in natural waters. Environ. Sci. Technol. 41,
2548e2553.
Sanderson, H., Johnson, D.J., Wilson, C.J., Brain, R.A., Solomon, K.R., 2003. Probabi-
listic hazard assessment of environmentally occurring pharmaceuticals toxicity
to fish, daphnids and algae by ECOSAR screening. Toxicol. Lett. 144, 383e395.
Siddick, Z.H., 2002. In: Malcolm, R. (Ed.), The Cancer Handbook, first ed. Alison John
Wiley & Sons, Ltd.
US EPA, 1993. Methods for Measuring the Acute Toxicity of Effluents and Receiving
Waters to Freshwater and Marine Organisms. EPA-600-4-90, fourth ed. US
Environmental Protection Agency, Washington DC.
Xie, H., 2012. Occurrence, ecotoxicology, and treatment of anticancer agents as
water contaminants. J. Environ. Anal. Toxicol. S 2.
Zhang, J., Tian, Q., Zhou, S., 2006. Clinical pharmacology of cyclophosphamide and
ifosfamide. Curr. Drug Ther. 1 (1), 55e84.
Zhang, Y., Xiao, Y., Zhang, J., Chang, V.W.C., Lim, T., 2017. Degradation of cyclo-
phosphamide and 5-fluorouracil in water using UV and UV/H2O2: kinetics
investigation, pathways and energetic analysis. J. Environ. Chem. Eng. 5 (1),
1133e1139.
Zhu, H., Long, M., Wu, J., Wang, M., Li, X., Shen, H., Xu, J., Zhou, L., Fang, Z., Luo, Y.,
Li, S., 2015. Ginseng alleviates cyclophosphamide-induced hepatotoxicity via
reversing disordered homeostasis of glutathione and bile acid. Sci. Rep. 17536,
1e14.
Zounkova, R., Odr aska, P., Dole , L., Hilscherov
zalova a, K., Mars ha, L., 2007.
alek, B., Bla
Ecotoxicity and genotoxicity assessment of cytostatic pharmaceuticals. Environ.
Toxicol. Chem. 26 (10), 2208e2214.