7-3-2021
The Environmental Fate
and Toxicity of Tyrosine
Kinase Inhibitors
Solving one problem might create another
Sedef Terzioğlu
Abstract
Pharmaceuticals have become indispensable to maintain a longer life and cure from diseases as cancer. In 2014, the World Health Organization (WHO) published stated an annual increase in cancer incidences, predicting to cause yearly 19.3 million cases by 2025, which number was already met in 2020 according to the International Agency for Research on Cancer. One subgroup of anti-cancer drugs is known as antineoplastics. They can pose an ecotoxicological risk since they are found as residues in aquatic environment from excretion through wastewater and are very potent drugs. One of the more recently developed group of antineoplastics are the Tyrosine Kinase Inhibitors (TKIs). Since the introduction of imatinib in 2001, 42 TKIs have been approved by the FDA and are currently used in the treatment of cancer. The rapid development of TKIs, however, comes with an increasing risk of pollution with a yet unknown effect on and fate in the environment.
Detection of 2 TKIs from aqueous environmental samples was performed using LC-MS/MS techniques. The limit of detection of imatinib and erlotinib in wastewater samples were 24-54 ng/L and 0.2-1.0 ng/L respectively. Both TKIs, had lower detection limits in human plasma samples. High resolution MS as a replacement of MS/MS with a quadrupole instrument has shown potential for a lower detection limit. The detection of other TKIs was only done for plasma samples, but most techniques used for these human samples also have potential to be used for environmental samples. The extraction of TKIs from aqueous samples still need more research since imatinib has shown a bad retention using most of the solid phase extraction cartridges. Next to that, the stability of erlotinib and imatinib in wastewater samples depend highly on the pH and temperature, which have to be considered.
Expected concentrations of 22 TKIs in surface water have been predicted using predicted environmental concentration (PEC) calculations. In European countries, imatinib was by far the most consumed TKI with the highest PEC of 7.99 ng/L in Belgium. In India, gefitinib was the most consumed TKI and gave PECs between 19.3 and 92.01, depending on the number of inhabitants taken for the calculations. The PECs are influenced by factors including the excretion rate and removal by the wastewater treatment plants, which both are factors that are not always available and can be assumed. Measured environmental concentrations (MECs) on the other hand, were only provided for erlotinib and imatinib. Both were measured at higher concentrations than their PECs, which gives rise to a possible underestimation of the predicted concentrations.
Toxicity studies were limited to the toxicity of imatinib as a representative for all TKIs on aqueous organisms and several plants. From all organisms, which include fish, rotifers, crustaceans, algae, cyanobacteria and bacteria, the lowest concentrations at which a significant effect was seen were the daphnids C. dubia. At a concentration of 430 ng/L, an EC10 was determined with inhibition of
reproduction as an endpoint after a 7-day chronic exposure. A LOAEC for its genotoxic effect on the C.
dubia was achieved at a concentration of 300 ng/L. Comparing these with measured and predicted
concentrations, a risk quotient could be measured using assessment factors for the toxicity. Using the PEC of imatinib a risk quotient of <1 was obtained, while this was 13.4 when using the highest MEC in literature, representing a high significant potential risk for adverse effects.
There is still a lack in knowledge of the fate of TKIs and the toxicities of different TKIs on the environment. Imatinib was taken as a representative, while studies on human health have shown that imatinib is not the most potentially toxic TKI brought to the market. Next to that, most TKIs are hydrophobic, and some are insoluble in water, which makes their partitioning to organic matter also interesting. In the future, research on the fate and their toxicity on organisms from a higher trophic level need attention, next to analytical methods to achieve detection at lower concentrations that meet the PEC concentrations.
Abbreviations
ALK Anaplastic lymphoma kinase
ALL Acute lymphoid leukemia
ATLD Alternating trilinear decomposition
BCR-ABL Breakpoint cluster region gene – Abelson
CBMN Cytokinesis-block micronucleus
CML Chronic myeloid leukemia
DAD Diode-array detection
DF Dilution factor
ECx Effect concentration (affects x%)
EDTA Ethylenediaminetetraacetic acid
EGFR Epidermal growth factor receptor
EMA European Medicines Agency
ESI Electrospray ionization
FDA Food and Drugs Administration
HPLC High pressure liquid chromatography
HRMS High resolution mass spectrometry
ICx Inhibitory concentrations (inhibits x%)
IDL Instrumental detection limit
LCx Lethal concentration (x% die)
LLOQ Lower limit of quantitation
LOAEC Lowest observed adverse effect concentration
LOD Limit of detection
LOEC Lowest observed effect concentration
MEC Measured environmental concentration
MRM Multiple reaction monitoring
NOAEC No observed adverse effect concentration
NOEC No observed effect concentration
NRTK Non-receptor tyrosine kinase
NSCLC Non-small cell lung cancer
OECD Organization for Economic Co-operation and Development
PEC Predicted environmental concentration
PNEC Predicted no effect concentration
QQQ Triple quadrupole
Q-trap Quadrupole-ion trap
RTK Receptor tyrosine kinase
SPE Solid phase extraction
SRM Single reaction monitoring
TDM Therapeutic drug monitoring
TKI Tyrosine kinase inhibitor
UPLC Ultra-performance liquid chromatography
VEGFR Vascular endothelial growth factor receptor
VFBIA Vibrio fischeri bioluminescence inhibition assay
WHO World Health Organization
WW Wastewater
Content
1 Introduction ... 4
2 Tyrosine Kinase Inhibitors ... 6
Tyrosine Kinase Activity ... 6
TKIs in Cancer therapy ... 6
2.2.1 TKI classes ... 7
2.2.2 Target of FDA-approved TKIs ... 7
3 Analysis of TKIs in human and environmental samples ... 8
Sample preparation ... 8
3.1.1 Sample Treatment ... 8
3.1.2 Sample Extraction ... 10
Detection and Quantification ... 11
3.2.1 LC and UV based detectors ... 12
3.2.2 LC-MS/MS (QQQ) ... 12
3.2.3 LC-HRMS ... 14
4 Environmental exposure ... 15
Predicted Environmental Concentrations ... 15
4.1.1 Method ... 15
4.1.2 Application ... 16
Measured Environmental Concentrations ... 18
5 Environmental Toxicity ... 20
Water organisms ... 23
5.1.1 Rotifers ... 23
5.1.2 Crustaceans/Daphnids ... 23
5.1.3 Algae & Bacteria ... 24
5.1.4 Zebrafish ... 25
Plants ... 26
6 Discussion ... 27
Treatment, detection and quantification of TKIs in environmental samples ... 27
Presence of TKIs in the environment ... 28
Comparison of PECs and MECs ... 28
Toxicity and environmental risk of TKIs ... 29
6.4.1 Imatinib as a representative for TKIs ... 29
6.4.2 Toxicities in-between organisms ... 30
6.4.3 Comparison of PECs and toxicity data (Risk) ... 30
Overall discussion ... 31
7 Conclusion ... 32
8 References ... 33
1 Introduction
Pharmaceuticals have a huge impact on human health. They have become indispensable to maintain a longer life and cure from diseases. Pharmaceuticals are designed to modify organic function causing a pharmacological effect within the human body and, through their biological activity, pose a potential risk to non-target organisms upon exposure. The development and usage of pharmaceuticals increases every year, resulting in an increase in environmental pollution as well. With the development of increasingly sensitive analytical methods, to detect very low concentrations of chemicals in a variety of matrices, the existence of pharmaceuticals in wastewater became a more recognized problem.1 In
contrast to release of pharmaceutical residues from the industries, residues from therapeutic use are released into the environment to a higher extend via urban or hospital wastewater after excretion. To some extent, pharmaceuticals remain unchanged or are metabolized into active compounds when leaving the body through feces and urine, which causes biologically active compounds to enter the environment. Studies on sewage treatment plants have shown to only partially remove pharmaceuticals from the wastewater.2 The environmental exposure and hazard by many classes of pharmaceuticals,
including oncological pharmaceuticals, on non-target organisms remains to be investigated.
In 2014, the World Health Organization (WHO) published a report stating the annual increase in cancer incidences and a predicted rise up to yearly 19.3 million cases by 2025.3 Figure 1 shows the
numbers of cancer cases and deaths worldwide in the year of 2020, determined by the International Agency for Research on Cancer (2020). Considering these numbers, the importance of developing and producing cancer related drugs rises correspondingly. One subgroup of anti-cancer drugs is known as antineoplastics. They are used for chemotherapy and are designed to prevent or interrupt rapid growth of tumor cells, affecting also non-target cells due to their low specificity (Heath et al. 2016).4 Combined
with a low therapeutic index, they can pose an ecotoxicological risk since they are found as residues in aquatic environment.5Antineoplastics enter the aquatic environment mainly by household effluents and
wastewater from hospitals and to a lesser extent from manufacturers.5,6 Biologically active compounds
that enter the aquatic system can lead to toxicity in, among others, aquatic organisms upon exposure. Knowledge on their environmental risk is scarce, and for some drugs less information is known than for others.
One of the more recently developed group of antineoplastics are the Tyrosine Kinase Inhibitors (TKIs). Since the introduction of imatinib in 2001, 42 TKIs have been approved by the FDA and are currently used in the treatment of cancer.7,8 The use of TKIs is limited by their sensitivity for acquiring
resistance which occur due to e.g. point mutations.9 Future treatment of resistant cancer variants with
The rapid development of TKIs, however, comes with an increasing risk of pollution with a yet unknown effect to the environment. Not only the effect on the environment and many non-target organisms is considered as an issue, also the side effects and possible toxicity that TKIs can bring to human health is still being investigated.11 Since TKIs for cancer treatment purposes are designed to
attack different pathways, e.g. the epidermal growth factor or vascular endothelial growth factor kinases, they might all give different side effects and toxicities. Their distribution in the environment is one of the important aspects to determine the exposure. Next to knowledge about their ecotoxicity, experiments with analytical methods to detect the drugs in different environmental compartments and biological samples in the best way possible are still being performed. Knowing the bioavailable concentrations and the concentrations at which the drug causes effects on an organism contributes to a risk assessment.
Providing an overview of these aspects, adds up to the knowledge of what has been done and what is still necessary to efficiently address the actual risks of TKIs and ways to prevent their exposure to non-target organisms. Awareness of the currently known toxicological effects is crucial to notice the problems in the environment. Therefore, this review will focus on current knowledge about the environmental toxicity of TKIs and discuss their environmental fate and analytical methods used for analysis.
2 Tyrosine Kinase Inhibitors
Tyrosine Kinase Activity
Tyrosine kinases are enzymes that play a role in the signaling of transduction processes within the cells. Activation of tyrosine kinases can for example lead to migration, cell proliferation and differentiation. It is a very important factor in the growth and reproduction of the cells in an organism. There are two different classifications of tyrosine kinase activity, which are the receptor tyrosine kinase (RTK) and the non-receptor tyrosine kinase (NRTK). In case of RTK, the activation takes place by extracellular binding of a ligand, while NRTK is caused by cytoplasmic proteins.12 There are
approximately 58 RTKs and 32 NRTKs identified.13
Tyrosine kinase enzymes are attached to the intracellular part of a receptor with a single transmembrane segment. Activation of the enzymes occur by connecting two receptors to form a dimer. As for RTK, this happens by the binding of ligands to the extracellular binding sites of a receptor. This can be done by means of one ligand with 2 binding sites, as can be seen in figure 2, or two separate ligands. The activation gives rise to phosphorylation on the intracellular receptor. Many different signaling proteins in the cell can now be activated by the phosphorus on the receptor. As for the NRTKs, the phosphorylation happens by more complex protein-protein interactions within the cell.12,14
Figure 2: Receptor Tyrosine Kinase Activity, reproduced from Alberts et al. 2014.
TKIs in Cancer therapy
Many tyrosine kinases are related to malignant tumors. Their activity can be used as a prognostic and predictive measurements of i.e. the aggressivity of the tumor or the response of treatment, by looking at overexpression or deregulation in cells by tyrosine kinases.15 The most investigated outcome of
tyrosine kinase activation within oncology is the growth factor signaling. This activity can convert into, among others, the proliferation of tumor cells and angiogenesis. Another activity playing an important role in oncology is the mutation of tyrosine kinases. Mutations of these enzymes cause an overexpression, meaning that the enzymes stay active and overstimulate i.e., the growth of tumor cells.16,17 TKIs are orally taken drugs designed for different target treatments. They are divided into
classes which say what site of the receptor they target. They can also be categorized by their target proteins, which are in some cases also related to a specific cancer.
2.2.1 TKI classes
TKIs have been developed for different tyrosine kinase activities, both RTK and NRTK. Some of the designed drugs target multiple signaling pathways, which means that they are not always specific for a single signaling process. The TKIs have been classified in types I to VI according to their binding sites on the receptor proteins.18,19 Types I and II TKIs bind to the active (I) and
non-active (II) ATP-pocket binding site. Types III and IV bind to allosteric binding sites next to the pocket (III) and on another side from the pocket (IV). Type V is the binding of a bivalent molecules that bind on two regions of the protein. Type VI, which is the most recently classified type, is the binding of drugs by covalent (irreversible) bonds on either the ATP-binding pocket, the allosteric binding site, or both. Figure 3 illustrated all these types.
2.2.2 Target of FDA-approved TKIs
The first TKI that was approved by the FDA is Imatinib, primary designed to target the BCR-ABL protein mutation in chronic myeloid leukemia (CML). BCR-BCR-ABL mutations cause an enhanced proliferation of myeloid cells and resistance to apoptosis of cells.20 Imatinib inhibits the phosphorylation
binding site, slowing down the proliferation and stimulating cell dead in the myeloid cell lines. Due to establishing resistance to imatinib after some time of treatment, more TKIs have been designed for the same main target protein.16 Imatinib is a multitarget drug and also inhibits other tyrosine kinases.
Inhibition of the epidermal growth factor receptor (EGFR) tyrosine kinase is important in the treatment of several different cancers. Six TKIs have inhibition of the EGFR as their main target, of which 5 have treatment of nonsmall-cell lung cancer (NSCLC) as their main target. Mutations in the EGFR leads to overexpression of proliferation of tumor cells, including the inhibition of cell death, metastasis and cell cycle progression. Mutations in the EGFR show the aggressiveness of a tumor.16
Several generation of EGFR inhibitors were brought to the market over the years. The first-generation drugs, erlotinib and gefitinib, were reversible type I inhibitors. The second generation included the drug afatinib, which is an irreversible type VI drug. This drug has shown major improvement in the inhibition of EGFR compared to the first generation. The third and for now last generation, osimertinib, has been developed to have a better compatibility toward mutated receptors.8
An increase in angiogenesis is caused by upregulation of the vascular endothelial growth factor (VEGFR).16 Angiogenesis, signifying the forming of new blood vessels, is important for the growth and
survival of tumor cells since these need a supply of oxygen.21 VEGFR inhibitors have treatment of solid
cancers as their main target. By inhibiting VEGFRs, angiogenesis can be down regulated. Treating VEGFRs and EGFRs together, has shown a synergistic effect.8
Next to the two major targets EGFRs and VEGFR, there are many other target tyrosine kinase proteins in cancer treatments with TKIs. In lung cancer treatment, the anaplastic lymphoma kinase (ALK) plays a role in the activation of several cell proliferation and cell survival pathways.22 Bruton
tyrosine kinase (BTK) is an important receptor in the signaling pathway of mantle cell lymphomas.23
TKI palbociclib is currently used a lot in the treatment of breast cancer by inhibition of the CDK4/6 (cyclin-dependent kinases) activity. B-Raf inhibitors are used in the treatment of BRAF melanoma, also known as an aggressive skin cancer. All other current TKIs with their main target protein kinase and Figure 3: An illustration of the classified TKI types I-VI, reproduced from Martinez et al. (2020)
3 Analysis of TKIs in human and environmental samples
Whether there is a case of pollution to the environment by TKIs can be determined by qualitative analytical methods. Since there is more information available on detection of TKIs in human samples, this could provide as a blueprint on how to detect TKIs from ecological samples. To acquire the concentrations at which TKIs are present in the environment after exposure and in patients after oral administration, suitable quantitative analytical methods are of great importance. The sensitivity, accuracy and precision of instruments play a big role in the correct determination of concentration. Next to that, a refined sample preparation is necessary for a good analysis.
Since there are important differences in human samples compared to environmental samples, the sample preparation will contain dissimilarities for both. On the other hand, from literature it is perceived that the analytical methods for detection and quantification of TKIs show many similarities for both sample types. Therefore, where sample preparation will be discussed for both sample types individually, the analytical methods used for both samples will be combined into one group. Another reason why human samples are taken into account in current study, is that many TKIs have not yet been analyzed using ecological samples at all. Most information is currently available for TKIs imatinib and erlotinib when it comes to detection methods in wastewater samples. Next to that, identification of TKIs from non-target animals go by means of blood sampling as well, for which sample preparation protocols of human blood samples can provide an initial sampling method.
Sample preparation
3.1.1 Sample Treatment
Human/other organisms
It is important to know how to store a biological sample to obtain a correct quantification of the target compound, without any degradation or metabolization between sample collection and measurement. Human samples that are being used for detection of TKIs are mostly plasma samples, and additionally serum, urine and dried blood stains. Many TKIs that have not been analyzed from environmental samples yet, have been extracted from blood samples already. For this particular reason, the stability of TKIs other than the already in environment treated compounds (imatinib and erlotinib) will be discussed.
Crizotinib in human and rat plasma samples have shown to be stable when stored at 4ºC (duration is not mentioned) and at room temperature for 24 hours in human plasma and for 5 hours in rat plasma.24 Gefitinib has shown a stability of 5 months at -20ºC and around 24 hours at a temperature
of 40ºC in a dried blood spot sample.25 Lenvatinib was stable in human serum and dried blood spot
samples at -20ºC for a month, at 4ºC for 3 weeks and for 24 hours at room temperature.26 Pazopanib,
lapatinib and sorafenib in human plasma were at least stable for 8 hours at room temperature and at least 21 days at -70ºC.27–29 Many TKIs in human EDTA plasma samples appeared to be stable at long term
storage (around 30 months) when stored at -20ºC and were stable up to 48 hours at room temperature. In these previous studies, plasma samples were mostly collected in EDTA or heparin containing tubes, mixed well and centrifuged, whereafter the plasma supernatant could be collected and stored frozen. Dries blood samples were obtained by dripping venous blood on a tissue paper and frozen.
Since plasma and serum samples are already vulnerable at room temperature and need to be used in a short time period, there is low chance that the TKIs lose stability when looking at stability data in literature. These stability studies give an indication at which temperature and for how long the TKIs could be stable in other, environmental, samples.
Environmental
The treatment of an environmental sample is of high importance for the correct quantification of a compound. From all environmental samples, most information was found on collecting TKIs from aquatic samples rather than air or soil samples. The concentrations in wastewater are low and need a good pre-treatment to maintain stability of the compounds in the sample. To be able to detect these low concentrations, degradation of the TKIs have to be prevented after collecting them. Drugs in wastewater samples can undergo processes in which they might metabolize or degrade, causing errors in the quantification. The reason that low concentrations are important to be detected is mostly because TKIs have a high potential toxicity because of their mechanism of action and can already cause toxicity at very low concentrations.
Negreira et al. (2014) has tested the stability of 26 different cytostatic drugs in wastewater, including the TKIs imatinib and erlotinib.30 Wastewater was collected from a large wastewater treatment
plant and stored in amber glass bottles in cold conditions. The wastewater samples were filtered and divided into 2 fractions, one fraction untreated with a pH of 7 and one fraction treated buffered with HCl to get a pH of 2. In some cases, acidification is necessary to prevent possible bacterial degradation, which is why the stability at a low pH is tested. Later on, we will see that acidification of imatinib is also contributes to a sufficient extraction from the samples. The samples are initially stored at -20ºC. To test the stability, the samples were individually spiked with the TKIs and stored at -20, 4 and 25 ºC respectively for 1 day up to 3 months. The results revealed that erlotinib and imatinib are relatively stable under all temperatures in a pH 7 solution during a storage period of 9 days. These compounds were more stable than many of the other cytostatic drugs that were tested. No clear difference was seen in acidic conditions within 9 days for erlotinib, while imatinib has shown a slight degradation at 4ºC storage for an unexplained reason. Over a longer period, both erlotinib and imatinib showed to be still relatively stable after 3 months of storage at -20ºC at a pH 7 condition. They did show, however, around 50% degradation after 3 months when stored at room temperature. While this article described the stability of erlotinib and imatinib in wastewater samples, the stability of other TKIs remains to be investigated.
The stability testing done on wastewater, was also performed before on HPLC water by Negreira et al. (2013b).31 Here, at a storage of -20ºC in neutral pH conditions, erlotinib and imatinib demonstrated
an approximately 70% response already after a storage time of 3 to 9 days. The effect of the storage time was more profound than in the wastewater samples, while both were treated in the same conditions. Because of this difference, eighter there might be a factor in the wastewater preventing degradation or a factor in the HPLC water that is stimulating degradation. Acidification of imatinib in an HPLC water sample gave a slightly worse result over a time period of 9 days, while erlotinib showed a slightly better response when acidified. Imatinib and erlotinib showed the same stability in HPLC water spiked with an organic compound when stored at 4ºC over 9 days.
Different studies in which the actual concentrations of cytostatic drugs were measured in environmental aqueous samples, have used the information from both studies of Negreira et al. (2013b and 2014) for their sample storage.30,32 For example, Isidori et al. (2016), stored their wastewater
samples at -20ºC for no longer than 2 months since this is the time period over which most cytostatic drugs were still stable.33 Since both wastewater and HPLC water are aqueous samples, it has to be
considered why there is a difference in stability of the compounds in both. This difference also provides us with an indication that there might be differences in stability of TKIs looking at samples from natural water collected from different areas. Because of this difference, it is recommended to use the natural water samples within 3 days from collection, taking possible degradation of 30% into account.
3.1.2 Sample Extraction
Human
Extraction of TKIs from biological samples most of the time varies from non-biological samples, mainly because biological samples contain different molecules such as proteins and peptides that have to be eliminated before the sample can be used for analysis. Human plasma samples used in TKI analysis were pre-treated with protein precipitation by Lankheet et al. (2013) Protein precipitation is a useful extraction method for blood samples to get rid of large protein molecules which are not favored to be present when inserting the sample in an LC column.34 Solid-phase extraction (SPE) was
used as well before for human and rat plasma samples.24 SPE is a pre-treatment where compounds in a
liquid sample are separated from the sample by means of their physical and chemical properties. In some studies, also no extraction method was used on the sample before inserting it into an LC column.
Environmental
Before analysis of the sample, the sample has to be extracted to get rid of interfering compounds. In literature, sample extraction is mostly done by SPE. Other ways of separations are known, such as liquid-liquid extractions and protein precipitation, but these will not be discussed since they have not been used yet for extraction of TKIs. Additionally, SPE is already a very suitable method for aqueous samples so no other methods are demanded till now.
Negreira et al. (2013a) performed sample extraction of 17 cytostatic drugs, including erlotinib and imatinib, from HPLC water samples using 5 different SPE cartridges.31 An on-line SPE-LC
instrument was used. The relative response of the chosen pharmaceutical drugs after extraction by the five different SPE cartridges are shown in figure 4. For all compounds, the HLB and PLRP-s cartridges resulted in the most preferred for extraction since they have more repeatability and more recovery than the others. Looking at only the TKIs, erlotinib shows an overall better recovery for most of the cartridges, while imatinib only showed sufficient recovery, though with low repeatability, when the PLRPs cartridge was used. The PLRPs cartridge, which is a crosslinked styrene divinylbenzene polymer, was chosen for further analysis of cytostatic drugs in collected samples from groundwater, river water and WWTP effluent and influent water. Before using this cartridge, a test was done on the recovery at different pH conditions, resulting in a better recovery at pH 2 rather than pH 6 for most of the samples.
Figure 4: The relative response of 17 pharmaceutical anti-cancer drugs after extraction by five different SPE cartridges, including TKIs imatinib (IMA) and erlotinib (ERL). The histogram is reproduced from Negreira et al. (2013a).
The performance of 4 different SPE cartridges for the extraction of several cytostatic drugs from wastewater was tested by Gómez-Canela et al. (2014).35 The tests were initially performed on Milli-Q
water samples that were spiked for 0.1 μg/L by the target cytostatic drugs. Between these drugs, again the TKIs imatinib and erlotinib were included in the analysis. From their testing, the best performance was achieved by the Oasis HLB cartridge, which has a polymeric reversed phase sorbent. In this study, the PLRPs cartridge was not used for comparison. The HLB cartridge did manage to significantly recover 24 from the 26 cytostatic drugs, but with imatinib being 1 of the 2 poorly/not recovered drugs. The HLB cartridge was further tested using wastewater samples collected from hospital WWTP effluents from 2 hospitals in North-East Spain and brought to 3 different pH conditions, namely 2, 3.5 and 7 pH. The best recovery was seen for the most acidic condition, where again imatinib did not show any recovery. Erlotinib on the other hand did show a good recovery. From the study of Gómez-Canela et al. (2014) using 2 different TKIs it can be concluded that in this case a positive sample extraction by a single SPE method is not acquired for 2 different TKIs.35 Development of a single sample extraction
method for all TKIs will need more attention in the future.
In addition, Santos et al. (2018) has tested more SPE cartridges to detect imatinib than was done in the studies before.36 After extraction, in most of the cartridges with different pH conditions imatinib
was not detected. However, the cartridges named Strata-XL-AW and Oasis Wax did show positive results extracting imatinib in 2 different pH conditions, namely pH 2 and pH 6. Because imatinib is a basic compound (pKa 8.1) it is positively charged in these conditions. The first cartridge was an anion cartridge, which is not the best solution to retain the positively charged imatinib, however, it did retain at very sufficient rates. This means that the reversed phase interactions within the sorbent were strong enough to make it stay in the sorbent. The reason why Santos et al. (2018) did not use a cation sorbent to retain Imatinib was because Imatinib was the only basic compound from all the cytostatic drugs they tested. In the future, using a cation cartridge might be the solution to extract imatinib, if no other pharmaceutical drugs are desired to be measured.
Detection and Quantification
Medical treatment with TKIs showed to have adverse effects on the human health. Due to these adverse effects, dose regulations are necessary to stay within the small therapeutic index of TKIs. In several studies a variability in concentrations were found in plasma samples of patients. Due to the high variability in patients and low therapeutic index, therapeutic drug monitoring (TDM) to determine the concentrations in patients is found to be promising. Mostly due to this particular reason, many research was done on the development of a method for detecting and quantifying TKIs from human samples.37
Compared to quantification methods in human samples, there is less research on quantification of TKIs in environmental samples. The most discussed environmental sample in which TKIs are found is hospital and domestic wastewater. Because of many similarities in separation and detection methods between human and environmental samples, studies on both were found useful to discuss. In many cases, for instance, more method development was done for a bigger variety of TKIs in human samples than for environmental samples.
In literature, LC-MS/MS is predominantly used for the determination of TKIs. There are variations seen in the LC method, such as use of HPLC or UPLC, and variations in the detection. Next to the most used detector, triple quadrupole MS/MS instrument in eighter SRM or MRM mode, other UV based detectors are also efficiently applied on human samples, but not yet on environmental samples.
3.2.1 LC and UV based detectors
UV detectors are less expensive and often more available in laboratories. Compared to MS detectors, UV based detectors are less sensitive, but easier to use. HPLC-UV based analysis were predominantly done for TDM analysis with human plasma samples. A reversed phase C18 column was used in all studies with changes in the mobile phase like the addition of a pH buffer, ratio of the organic and aqueous mobile phase solvents and the use of an isocratic or gradient mobile phase. Most TKIs can be detected at wavelengths between 260 and 265 nm. The lower limit of quantitation (LLOQ) was different for every TKI, with the lowest value of 5 ng/mL for ponatinib and the highest value of 500 ng/mL for pazopanib.38
Another UV based detection method, diode-array detection (DAD), was used by Fouad et al. (2015) and Xiang et al. (2015).37,39 The methods were developed and optimized to determine
concentrations of the dasatinib, pazopanib, vandetanib, afatinib and ibrutinib in plasma samples for the purpose of TDM. Plasma samples were gathered and separately spiked with the TKIs. As for the chromatographic features, a reversed phase C18 column was used in both cases, but with different dimensions corresponding to UPLC and HPLC. The mobile phase was also different in both studies. Where Fouad et al. (2015) used a gradient mobile phase containing an ammonium formate buffer and an increasing percentage of acetonitrile, from 5%-90%, Xiang et al. (2015) optimized an isocratic mobile phase of 65% methanol and 35% water.
In both studies, the compounds were separated within an analysis time of 4 minutes. For all compounds, an LLOQ of 5-70 ng/mL in plasma were found. The TKIs dasatinib, afatinib, pazopanib and vandetanib were also separated and detected from a mixture solution containing all 4 TKIs. There was an overlap seen in the chromatograms between the compounds or between compound and interferences, resulting in an incomplete separation of the TKIs. Despite the insufficient separation, they could get quantitative information about the targeted analytes by means of an alternating trilinear decomposition (ATLD) algorithm. Both studies have shown that UPLC-DAD is useful for quantitation of 5 different TKIs in plasma samples.
3.2.2 LC-MS/MS (QQQ)
Separation of TKIs in human samples is mostly done by HPLC, but other than the UV-based detectors, there is a huge research field on detecting the TKIs by MS techniques. Sabourian et al. (2020) have written a review on many HPLC methods to detect and quantify anticancer drugs.38 Many of the
TKIs could be detected by making use of mostly tandem MS methods. As an example, Van Veelen et al. (2019) developed and validated a method to detect osimertinib in human plasma by HPLC-MS/MS.40
They made use of a triple quadrupole MS (QQQ) in the multiple reaction monitoring (MRM) mode and managed to get a precision and accuracy according to the guidelines using this technique. Moreno et al. (2013) used a triple quadrupole MS in single reaction monitoring (SRM) mode to make a method to detect imatinib in plasma samples.41 Other examples of TKIs detected with eighter SRM or MRM mode
MS/MS are crizotinib, lenvatinib and vemurafenib.24,26,42 Because TKIs are always taken together with
other cancer drugs, detecting them in mixtures is also an interesting subject. Lankheet et al. (2013) have shown that it is possible to quantify 8 different TKIs mixed in a human plasma sample by triple quadrupole MS/MS in MRM mode.34 Another study by Mukai et al. (2020) revealed that it is also
Table 1: Methods for the detection and quantification of imatinib and erlotinib in (waste)water samples.
TKI Sample Method LOD
(ng/L) LLOQ (ng/L) Literature Imatinib Groundwater RP LC-MS/MS using triple quadrupole in SRM mode. 24 ng/L 80 ng/L Negreira et al. (2013a) Surface water 45 ng/L 150 ng/L WWTP eff. 36 ng/L 120 ng/L WWTP inf. 54 ng/L 180 ng/L HPLC water 22 ng/L 75 ng/L
HPLC water RP UPLC-MS/MS using triple quadrupole in SRM mode. 8 ng/mL 26.7 ng/mL Negreira et al. (2013b) HPLC Water RP HPLC-MS/MS using triple quadrupole in SRM mode. 2.5 ng - Gómez-Canela et al. (2013a) HPLC & Wastewater RP HPLC-HRMS with an Orbitrap 0.25 ng - Gómez-Canela et al. (2014) Erlotinib Groundwater Reversed phase LC-MS/MS using triple quadrupole in SRM mode. 0.2 ng/L 0.5 Negreira et al. (2013a) Surface water 0.7 ng/L 2.3 WWTP eff. 1.0 ng/L 3.4 WWTP inf. 0.5 ng/L 1.7 HPLC Water 0.1 ng/L 0.3
HPLC water Reversed phase
UPLC-MS/MS using triple quadrupole in SRM mode. 0.1 ng/mL 1 ng/L Negreira et al. (2013b) HPLC Water RP HPLC-MS/MS using triple quadrupole in SRM mode. 0.01 ng - Gómez-Canela et al. (2013) HPLC & Wastewater RP HPLC-HRMS with an Orbitrap <0.005 ng - Gómez-Canela et al. (2013 & 2014)
More interesting for current review is the detection in environmental samples. In table 1 the studies recently done for detecting the TKIs imatinib and erlotinib in (waste)water samples with LC-MS/MS techniques are demonstrated. For the detection of erlotinib and imatinib in pure water samples, Negreira et al. (2013b)32 developed an analytical
method using a UPLC, with a reversed phase C18 column, for separations connected to a triple quadrupole MS/MS detector with an electrospray ionization (ESI) source in positive mode. They performed experiments on single compound samples and 24 cytostatic drugs mixed together, all in SRM mode using parameters that are optimized for each compound. A gradient mobile phase was used with ultrapure water (solvent A), methanol (solvent B) and the addition of a 0.1% formic acid buffer. The gradient
mounted up from 5% methanol at 0 min to 100% methanol at 23 minutes. In their next research, they used this same method to detect erlotinib and imatinib in spiked wastewater, obtaining approximately the same quality in their results.30
Negreira et al. (2013a) also used LC-MS/MS to detect imatinib and erlotinib in spiked groundwater, surface water and the effluent and influent of WWTP.31 In this case, regular LC was used.
The limit of detection (LOD) for imatinib is sufficiently higher than that of erlotinib due to inefficient ionization of imatinib in the ESI interface. Next to that, they looked at possible matrix effects and found out that for both TKIs there is a potential suppression for surface water and wastewater samples, as expressed in figure 5. This same method has been further used in other studies as a standard to detect these samples from wastewater for purposes of determining the exact measured concentrations.
3.2.3 LC-HRMS
Gómez-Canela et al. (2013) used LC-MS/MS in the same chromatographic conditions as Negreira et al. (2013a), with a slight difference in the chromatogram gradient time, and compared this to LC-HRMS (high resolution MS) with an orbitrap instrument instead of the less sensitive triple quadrupole.44 The chromatographic conditions were kept the same. As for the parameters of the
Orbitrap, a full scan data acquisition was used over a small molecule size range. Improved fragmentation was achieved through higher energy collision dissociation voltages. Both methods showed similar selectivity for detection of imatinib and erlotinib. For some of the other measured cytostatic drugs, using HRMS provided better sensitivity than using the triple quadrupole MS. In their case the instrumental detection limit (IDL) for imatinib was 2.5 ng and for erlotinib was 0.01 ng, using the quadrupole. For erlotinib the IDL was lower when the HRMS was used, while for imatinib no IDL was detected. In a follow up study by Gómez-Canela et al. (2014), the LC-HRMS method was used on spiked wastewater samples as well providing similar results.35
Figure 5: Matrix effects in groundwater, surface water and wastewater influent and effluent. Reproduced from Negreira et al. (2013b)
4 Environmental exposure
Predicted Environmental Concentrations
The predicted effect concentration (PEC) gives information about how much of a chemical is expected to be available in the environment. Since there is less known about actually measured concentrations, the PEC can be calculated and used to assess the environmental risk of compounds. Combined with the predicted no effect concentration (PNEC), determined from toxicity testing, a risk assessment can be made for a certain chemical compound.
4.1.1 Method
To predict concentrations of pharmaceutical drugs in the environment, a guideline was made by the European Medicines Agency (EMA), with the most recent revision in 2018.45 As for PEC
calculations from wastewater or surface water samples an equation based on this guideline is used by Cristóvão et al. (2020).46 The PEC for surface water, river water and wastewater could be calculated by
Eq. 1:
𝑃𝐸𝐶 =!"#$%&'()"# ∗ ,!"# ∗ (./,$$%&)
11'()*+ ∗ )#234 ∗ 5, (1)
Consumption refers to the consumed pharmaceuticals in kg/year; Fexc is the fraction excreted from the body by urine and/or feces as the parent
compound; FWWTP is the fraction of the parent that is removed by a WWTP; WWinhab is the volume of wastewater that is produced by inhabitants
per liter per year; inhab is the number of inhabitants; DF refers to the dilution factor and is the dilution from the WWTP to the surface water. When calculating PEC for the influent and effluent of a WWTP, there is no DF.
Whether the calculated PEC is close to the actual concentrations in the environment, depends highly on the available data. The consumption of pharmaceutical drugs varies between countries. Variations can be seen in, for example, the quantity of elder population and developed countries compared to under-developed countries. Consumption is known to be higher in countries where a better health system is available, and a higher rate of seniors is living. Next to the consumption, the number inhabitants per certain area contributes to a higher consumption. A populous area with a higher rate of elders, give higher consumption. Consumption data has to come from a reliable source and in most cases, data are collected from pharmacies and hospitals. The consumption and inhabitants are parameters that depend on the country, just as the volume of wastewater produced per inhabitant and the dilution factor. The EMA has set a default for the dilution factor of 10, but for some countries, a better calculated value is available in literature.47 It has to be considered that the dilution factor, which is a parameter for
how much of the effluent water is diluted to the surface water or river water, can change the PEC in factors over 100.48
Other parameters are important specifically for the pharmaceutical drug that enters the waters by human excretion. Factors influencing the amount of the pharmaceuticals to enter the sewerage are the fractions of a parent compound to be excreted by the urine and feces unchanged. Since in many developed countries the sewage system includes a wastewater treatment plant (WWTP) the fraction that is cleaned by the WWTP needs to be taken into account as well. Data for both, Fexc and FWWTP, are not
always available. In most cases, when no data is available, both are set to a fraction of 0.5 assuming that there is a certain amount of excretion and cleaning. These fractions can have a huge negative effect on the precision of the PEC when they are not known and are also known to affect the PEC most significantly compared to the other parameters in Eq 1.48 The excretion rate is mainly depended on the
metabolic activity inside the human body, while the rate of removal by the WWTP is dependent on the hydrophobicity and persistency of the pharmaceuticals. The FWWTP can be predicted by the KOW or KOC
4.1.2 Application
For several TKIs, PECs of surface water are calculated with data from a variety of European countries and for India. In all cases, the PEC was calculated using the same parameters as in Eq 1. An overview of PECs in surface/river waters of 22 different TKIs can be found in table 2. In all of the studies, a PEC was calculated for imatinib and erlotinib, while in some studies more TKIs were included in the calculations than in others.
Table 2: An overview of all PECs calculated for 22 different TKIs in surface or river water from a variety of studies and countries. All PECs are displayed in ng/L.
a Cristóvão et al. (2020) b Santos et al. (2017) c Besse et al. (2012) d Booker et al (2014) e Franquet-Griell et al. (2015)
Data from France was provided for 4 TKIs by Besse et al. (2012).50 National consumption data
was obtained by a health products agency and local data from regional care centers, resulting in an overall consumption data for whole France from the years 2004 and 2008. Since consumption of TKIs is higher each year and since more TKIs were in the market in 2008, only data from 2008 was discussed here. Default values were taken for the DF (10) and the WWinhab (200 L per person per day). In their
analysis, no removal rate from the WWTP was taken into account because there was no information available for the removal rates of the TKIs that time. Therefore, the fraction removed from the water was set to 0 and causes a higher value than when the TKIs are actually (partly) removed by the WWTP.
TKI Lisbon, Portugal (2016)a Portugal (2007-2015)b Belgium (2015)a France (2008)c NW England (2010-2012)d Catalonia Spain (2012)e India 3 states (2016)a India 1 state (2016)a Axitinib 0,00407 Bosutinib 0,0486 0,14 Ceritinib 0,00478 Cobimetinib 0,00381 Crizotinib 0,0689 0,36 0,237 Dabrafenib 0,0224 Dasatinib 0,107 0,15 0,294 0,05 Erlotinib 0,00736 0,06 0,0537 0,07 0,1 0,02 0,0535 0,256 Gefitinib 0,608 0,08 2,09 0,02 19,3 92,1 Ibrutinib 0,651 0,17 Imatinib 1,86 5,37 7,99 4,99 0,5 2,34 2,77 13,3 Lapatinib 0,033 0,47 0,793 1,86 0,26 Nilotinib 0,148 0,41 0,00893 0,8 0,31 Osimertinib 0,0000772 Pazopanib 0,637 1,9 6,04 0,08 Regorafenib 0,000781 Ruxolitinib 0,00175 Sorafenib 0,0971 0,27 0,405 0,2 0,848 4,06 Sunitinib 0,0698 0,11 0,31 0,02 Trametinib 0,00146 Vandetanib 0,0205 Vemurafenib 0,697 0,9 0,266
2004, where a value of 3.33 ng/L was obtained. The reason for this is that in 2004 Imatinib was one of the few TKIs that was already available on the market and consumed in a rapidly increasing rate. Between 2004 and 2008 an increase in the PEC for imatinib is observed.
PECs in waters in North-West England were calculated by Booker et al. (2014) according to data from 2010 to 2012.51 The most remarkable about the data that was used by Booker et al. (2014), is
that body excretion rates were limited to urinary excretion only. This causes huge differences compared to the PECs calculated in other studies. For example, a great difference is seen for TKI lapatinib in the Fexc used by Booker et al, (2014) of 0.01 and the Fexc used by Besse et al. (2012) of 0.7. It has to be taken
into account that Besse et al. (2012) took the highest Fexc through feces known from the BCB
Micromedex database. This database demonstrated a median of 27% with the highest excretion at 67% for lapatinib through feces only. Lapatinib is an example of the TKIs that have a high excretion rate through the feces. These differences between both studies indicate that the excretion through both urine and feces have to be taken into account. The PEC calculated for lapatinib by Booker et al. (2014) was 0.00 ng/L and therefore not added to the data in table X. Booker also used default values for the DF (10) and WWinhab (200 L per person per day).
Franquet-Griell et al. (2015) have calculated the PEC for several of the most used cytostatic drugs in 2012 in Catalonia, Spain.52 A different default value was set for the WW
inhab, which was in this
case 139 L per person per day for inhabitants of Catalonia. A lower value for WWinhab gives rise to the
PEC on the contrary to a higher WWinhab. Franquet-Griell et al. (2015) have chosen to use an estimated
value for the DF which was calculated using geographic information by a study of Keller et al. (2014).47
A DF of 25.92 was chosen as an average of dilution in whole Spain. In this study, also the values collected from previous studies were recalculated using estimated DFs of 75.73 for France and 37.16 for the UK, concluding in a lower PEC for both countries. These values, however, were chosen not to be displayed in current study since these are recalculations which were not done by the same authors as the main study. Also, compared to all other studies where PECs are calculated for TKIs, only in this case the DF value deviates from the default which was set by the EMA guidelines for Europe.
In two studies, data from Portugal was used to determine the PECs in surface waters of whole Portugal by Santos et al. (2017) and particularly in Lisbon by Cristóvão et al. (2020).46,53 Santos et al. (2017) has calculated PECs for 5
different regions in Portugal, using specific data for each region, coming to a total PEC for whole Portugal at the end, all with data from 2007 to 2015. The regions with their consumption data and water use are shown in figure 6. Here it is seen that the inhabitants and water consumption vary much between regions. The same was true for the anticancer drugs consumption. In the 2 southern regions, Alentejo and Algarve, lesser volumes were consumed of, for example imatinib. While imatinib was consumed in volumes of 2,06 kg/year in Algarve, 56,2 kg/year was consumed in Lisbon. These values show that there is a lot of within country variation that could influence the actual risk calculations which are done together with the toxicity knowledge. This is due to the fact that the PEC of imatinib for
whole Portugal, 5.37 ng/L, is not representative for Algarve, for which a PEC of 0.94 ng/L was obtained. Cristóvão et al. (2020) focused on calculating PECs for Lisbon only in 2016. There is a big difference between the calculated PEC for, among others, imatinib. As mentioned before, Santos et al. Figure 6: Five regions in Portugal with their corresponding number of inhabitants and water consumption. The figure is reproduced from Santos et al. (2017)
expected. It is, however, noticed that the information on consumption data of Cristóvão et al. (2020) was a lot less than for Santos et al. (2017), which was probably due to the fact that Cristóvão et al. (2020) obtained data from only one cancer institute while Santos et al. (2017) had information from all hospitals and pharmacies in that region. In the other used parameters, only small differences in the inhabitants and water used per inhabitant were found.
Cristóvão et al. (2020) obtained data for consumption of antineoplastics only from a national health institute in Belgium.46 Taking imatinib as an example, a total consumption of 165 kg a year was
achieved from the data for Belgium in 2015. Compared to data from Portugal in 2015, which was 148 kg a year, more imatinib was consumed in Belgium.53 From both data, Belgium has approximately 1,5
million more inhabitants, but do show a lower water use in general. These consumption details, contribute to the higher PEC for Belgium, as seen in table 2.
The data from India, was obtained from one cancer hospital in the state Haryana. This hospital treats patients from 3 different states. For this particular reason, the PECs were calculated for the population of 1 state and for that of 3 states. Compared to European countries, gefitinib is by far the most used TKI in India, while in Europe imatinib is predominantly used. In all of the studies mentioned before, the excretion of gefitinib particularly through the feces was not taken into account, possibly because of the lack information back then. The excretion through urine only is not very high for gefitinib, only 5%, while excretion through feces is approximately 86% according to data from Cristóvão et al. (2020). The excretion rate explains for the majority why the PECs for gefitinib in India are very high compared to what has been calculated before.
Measured Environmental Concentrations
At the time of writing, most studies on the measured environmental concentrations (MEC) of cytostatic drugs in water samples were limited to TKIs erlotinib and imatinib. The majority of environmental concentration data was collected in Spain, and less research is done in other countries. Measuring imatinib in water samples has shown to be less sensitive than erlotinib, possibly due to the fact that the LOD of imatinib is much higher using LC-MS/MS techniques. In some cases, imatinib could be found in wastewater above the LOD but no exact concentration could be measured since the LLOQ was not exceeded.
Using UPLC-HRMS, Gómez-Canela et al. (2014) tried to detect erlotinib in wastewater collected from before and after a WWTP of 2 big hospitals in North-East Spain.35 From all 19 cytostatic
drugs they only detected 7 where erlotinib was not one of the detected drugs. Additionally, in a later study of Franquet-Griell et al. (2017) erlotinib was successfully detected in surface water.54 The
wastewater samples were collected from 19 different spots in the Bèsos River. To avoid stagnant water, samples were taken from 20 cm below the actual surface. The Bèsos River was chosen because it is located at a highly industrial and populated area with the estuary located in the big city of Barcelona. In only one sample from a sampling spots far from the estuary, erlotinib was detected, with a concentration of 3.9 ng/L. The site where this was detected had a lower flow rate than other sites on the river. The MEC of 3.9 ng/L was therefore higher than what they calculated as their PEC, which was <0.1 ng/L, because a higher flow rate was used in the calculations.
Imatinib was measured in concentrations above the LOD and below the LLOQ in hospital wastewater samples in Ljubljana.33 In their case, the LOD of imatinib was between 36 and 54 ng/L and
the LLOQ between 120 and 180 ng/L, depending on the water sample. This means that the concentration of imatinib detected in hospital wastewater from Ljubljana must be approximately between 36 and 180 ng/L. In the same study by Isidori et al. (2016), erlotinib concentrations were also measured in hospital wastewater from Ljubljana and from Barcelona.33 In both countries, concentrations for erlotinib of
from 2 different areas of a hospital in Valencia. From the first area, nuclear medicines enter the water while from the second area, the daily use of the hospital waste enters the wastewater.55 In the wastewater
sample taken from where most nuclear drugs are discharged, imatinib was measured in quantities of 75-577 ng/L, while in the other sample, no imatinib was detected. Considering these results, the area of sampling is found particularly important.
In another study, analysis of imatinib was implemented on environmental samples collected from a catchment in South West England.56 Next to three water samples (surface water, influent and
effluent of a WWTP), also solid particulate matter and digested solids were collected. Since imatinib has a KOW of approximately 3, it is expected that the compound will be available in higher concentrations
in organic matter at equilibrium. In the water samples, concentrations of ±38.3 ng/L in surface water and ±143.3 ng/L in effluents were found, while ±47.6 ng/L was found in solid particulate matter and ±123.0 ng/g in digested solids. The WWTP influent concentrations were lower than the effluent concentrations. As a possible reason for this phenomenon, Proctor et al. (2019) said that conjugated metabolites could turn back to their parent molecule.56 This is not necessarily the case for TKIs, since
5 Environmental Toxicity
Because of the mechanism of action, anticancer drugs have a high potency to form toxicity to organisms. Due to the fact that TKIs are not very selective, they also attack healthy, non-cancer cells which is why their effect on organisms is of high importance. Together with the phenomenon that the usage of these drugs is increasing each year, they can form a risk to exposed non-target organisms.46,57,58
The first TKI brought to the market in 2001, imatinib, is currently the most researched TKI. Imatinib is still being consumed a lot and therefore the PEC is in many cases higher than the other TKIs. When looking at the toxicity of TKIs, toxicity testing on aquatic organisms is almost exclusively done using imatinib as a representative for the TKI group. In literature, few data were found on toxicity testing of other TKIs on aquatic organisms. In this review, the focus is set on toxicity testing of imatinib on water organisms and plants, and on acute toxicity testing of a few other TKIs on zebrafish. All data found on ecotoxicity testing for imatinib only can be read from table 3 for each taxon respectively.
Together with the PEC, toxicity data takes part in the calculation of the potential risk of a drug to the environment. There is no risk without hazard and no risk without actual exposure. This risk is expressed in the risk quotient (RQ), which can be calculated as shown in eq. 2. As for the interpretation, A potential risk to the environment is given when RQ is higher than 1.59 The PNEC is a prediction from
toxicity data which can be defined from single toxicity data multiplied by a necessary assessment factors for the uncertainty of the data, or by species sensitivity distributions (SSD). The latter was not done before for any of the TKIs, which makes the first one the only currently available method for assessing an RQ. Assessment factors are factors of usually 10x from which the toxicity measurement is divided.
If there is less toxicity data available on fewer species from different trophic levels and if there is no chronic toxicity data available, the highest assessment factor is used and is a factor of 103 in most cases.
The assessment factors can be found in several guidelines.
𝑅𝑄 = 678
6978 (2)
Interpretations for aquatic organisms is as followed: No significant risk at RQ < 1.0; Small potential risk for adverse effects at 1.0 ≥ RQ < 10; high significant potential risk for adverse effects for 10 ≥ RQ < 100; Potential risk is expected at ≥100.59
Table 3: An overview of all environmental toxicity data provided from literature for imatinib.
Species/Genus Parameter C in mg/L Exposure Endpoint Literature
Rotifers
B. caliciflorus LC50 3,82 24h acute Lethality Parrella et al.
(2014a)
EC50 0,74 48h chronic Population growth
inhibition EC20 0,26 EC10 0,14 NOEC 0,07 LOEC 0,15 Crustaceans
T. platyurus LC50 43,27 24h acute Lethality Parrella et al.
(2014a)
C. dubia LC50 31,92
EC50 0,115 7d chronic Inhibition of
reproduction EC20 0,003 EC10 0,00043 NOEC 0,00027 LOEC 0,00087 NOAEC 0,00003 24h comet assay
Genotoxicity Parrella et al. (2015)
LOAEC 0,0003
D. magna EC50 72,43 48h acute Immobility Białk-Bielińska et al.
(2017)
EC50 11,97 Parrella et al.
(2014a)
EC50 0,308 21d chronic Inhibition of
reproduction EC20 0,03162 EC10 0,00834 NOEC 0,00298 LOEC 0,00954 NOAEC 0,0002 24h comet assay
Genotoxicity Parrella et al. (2015)
LOAEC 0,002
Bacterium
V. fischeri EC50 26,06 30m acute Bioluminescence Białk-Bielińska et al. (2017)
Cyanobacterium
S. leopoliensis EC50 5,36 72h acute/
chronic bioassays Population growth inhibition Brezovšek et al. (2014) EC20 3,81 EC10 3,12 NOEC 3,84 LOEC 4,00
Table 3 (continued): An overview of all environmental toxicity data provided from literature for imatinib.
Algae
R. subcapitata EC50 5,08 72h acute/
chronic bioassays
Population growth
inhibition Białk-Bielińska et al. (2017)
EC50 2,29 Brezovšek et al.
(2014) EC20 1,17 EC10 0,79 NOEC 0,38 LOEC 1,19 Zebrafish D. rerio (adult)
LC50 70,8 96h acute Lethality Kovács et al. (2016)
D. rerio (juveniles) LOEC 10 33d chronic NOEC 1 D. rerio (embryo) LC50 158,3 48h acute Lethality, embryonic malformation & hatching delay LC50 141,6 72h acute LC50 118,0 96h acute LC50 65,9 120h acute D. rerio (liver cells)
IC50 21,83 4h Cytotoxicity Novak et al. (2017)
IC50 9,1 24h IC50 6,23 48h IC50 4,04 72h LOEC 1 24h Genotoxicity - Comet assay NOEC 0,1 LOEC 0,01 72h NOEC 0,001 LOEC 1 72h Genotoxicity - CBMN assay NOEC 0,1 Plants
L. minor EC50 61,05 7 days Growth inhibition Białk-Bielińska et al.
(2017)
L. sativa (seeds)
IC25 46,69 Unknown Root Growth Pichler et al. (2014)
IC10 16,01
Allium LOEC 0,6 24-72h Induction of
micronuclei
NOEC 0,06
Tradescantia LOEC 5,9 Pichler et al. (2014)
& Mišík et al. (2016)
Water organisms
5.1.1 Rotifers
Rotifers have an impact on the ecosystem since they contribute to a clean ocean by nutrient recycling and are active in converting food for higher trophic levels. Next to their important role, they are also very attractive for ecotoxicity testing because their small size, short lifespan (4-5 days), fast growth (0.5 to 1.5 days) and fast reproduction. What needs to be considered when using rotifers, is their great diversity for being susceptible to toxicants between species.60
Acute and chronic toxicity testing on Brachionus calyciflorus are performed looking at lethality, growth inhibition and feeding behavior as endpoints.61,62 Acute testing by Parrella et al. (2014a) was
performed in a static condition, exposing rotifers of <2 hours old for 24 hours with lethality as the endpoint. Resulting from the data analysis, they have achieved an LC50 of 3.82 mg/L. Yan et al. (2017)
resulted in an LC50 of 11.743 mg/L for a 24-hours exposure time. Chronic testing was performed for a
time period of 48 hours. From the derived concentration/effect curves, they achieved an EC50, EC20,
EC10, NOEC and LOEC of 740, 260, 140, 70 and 150 µg/L respectively.61 Studying the feeding behavior
of rotifers when exposed to imatinib, has shown feeding depression in the rotifers. Observing their filtration and ingestion rates has shown that at even low concentrations (minimum of 0.12 µg/L) there was a significant impact on the feeding of rotifers after 12h post-exposure and 12h re-exposures. The re-exposure was done to see if there is any recovery between 2 exposure periods. It was noticed that the feeding depression was less intense for the re-exposure period, meaning that the rotifers show a slight adaption to the imatinib.62 There were no effect concentrations or no observed effect concentrations
calculated for the feeding behavior.
5.1.2 Crustaceans/Daphnids
Crustaceans play an important role in the food chain. Large crustaceans, like crabs, are used as food for predators higher up in the food chain, while small crustaceans like daphnids help in nutrient recycling. Daphnids are used a lot in toxicity testing due to their sensitivity and convenient life cycle. Also, daphnids are suitable for testing endpoints other than lethality because of their appearance.
Acute and chronic toxicity testing on daphnids Daphnia magna and Ceriodaphnia dubia, and acute toxicity testing on fairy shrimp Thamnocephalus platyurus were performed.61,63 The fairy shrimp
was selected because of its high sensitivity to toxicants and both daphnids because they are widely used for toxicity testing. As for the acute toxicity testing, both T. platyurus and C. dubia were exposed to imatinib over a 24-hour time period, measuring lethality as endpoint. D. magna were exposed to imatinib over a 48-hour time period with the effect on immobility as an endpoint in 2 different studies. As for the chronic testing, the inhibition on reproduction was tested for both daphnids, where C. dubia were exposed for 7 days and D. magna for 21 days.
Despite the reason that T. platyurus is chosen for its sensitivity to toxicants, the lowest toxicity between the three crustaceans was obtained for this species with an LC50 of 43.27 mg/L. The acute
toxicity testing on the C. dubia provided an LC50 of 31.92 mg/L. As for the D. magna, which was tested
in 2 different studies, an acute EC50 of 11.97 mg/L was obtained in the first study while an EC50 of 72.43
mg/L was found in the second study.61,63 Looking at chronic toxicities, imatinib has shown to have the
biggest effect on the C. dubia with an EC10 of 0.43 µg/L and a NOEC of 0.27 µg/L. As for the D. magna,
the calculated toxicities were 308, 31.62, 8.34, 2.98 and 9.54 µg/L for the EC50, EC20, EC10, NOEC and
LOEC respectively. Parrella et al. (2014b) also tested mixture toxicity of imatinib in combination with other cancer drugs on both C. dubia and D. magna.64 Mixing imatinib with cancer drugs cisplatin and
