OECD/OCDE 241 Adopted: 28 July 2015 OECD GUIDELINE FOR THE TESTING OF CHEMICALS The Larval Amphibian Growth and Development Assay (LAGDA) INTRODUCTION 1. The need to develop and validate an assay capable of identifying and characterizing the adverse consequences of exposure to toxic chemicals in amphibians, originates from concerns that environmental levels of chemicals may cause adverse effects in both humans and wildlife. The test guideline of the Larval Amphibian Growth and Development Assay (LAGDA) describes a toxicity test with an amphibian species that considers growth and development from fertilization through the early juvenile period. It is an assay (typically 16 weeks) that assesses early development, metamorphosis, survival, growth, and partial reproductive maturation. It also enables measurement of a suite of other endpoints that allows for diagnostic evaluation of suspected endocrine disrupting chemicals (EDCs) or other types of developmental and reproductive toxicants. The method described in this guideline is derived from validation work on African clawed frog (Xenopus laevis) by the U.S. Environmental Protection Agency (U.S. EPA) with supporting work in Japan (1). Although other amphibian species may be adapted to a growth and developmental test protocol with ability to determine genetic sex being an important component, the specific methods and observational endpoints detailed in this guideline are applicable to Xenopus laevis alone. 2. The LAGDA serves as a higher tier test with an amphibian for collecting more comprehensive concentration-response information on adverse effects suitable for use in hazard identification and characterization, and in ecological risk assessment. The assay fits at Level 4 of the OECD Conceptual Framework on Endocrine Disrupters Testing and Assessment, where in vivo assays also provide data on adverse effects on endocrine relevant endpoints (2). The general experimental design entails exposing X. laevis embryos at Nieuwkoop and Faber (NF) stage 8-10 (3) to a minimum of four different concentrations of test chemical (generally spaced at not less than half-logarithmic intervals) and control(s) until 10 weeks after the median time to NF stage 62 in the control, with one interim sub-sample at NF stage 62 (≤ 45 post fertilization; usually around 45 days (dpf). There are four replicates in each test concentration with eight replicates for the control. Endpoints evaluated during the course of the exposure (at the interim sub-sample and final sample at completion of the test) include those indicative of generalized toxicity: mortality, abnormal behavior, and growth determinations (length and weight), as well as endpoints designed to characterize specific endocrine toxicity modes of action targeting oestrogen, androgen or thyroid-mediated physiological processes. The method gives primary emphasis to potential population relevant effects (namely, adverse impacts on survival, development, growth and reproductive development) for the 1 © OECD, (2015) You are free to use this material for personal, non-commercial purposes without seeking prior consent from the OECD, provided the source is duly mentioned. Any commercial use of this material is subject to written permission from the OECD. 241 OECD/OCDE calculation of a No Observed Effect Concentration (NOEC) or an Effect Concentration causing x% change (ECx) in the endpoint measured. Although it should be noted that ECx approaches are rarely suitable for large studies of this type where increasing the number of test concentrations to allow for determination of the desired ECx may be impractical. It should also be noted that the method does not cover the reproductive phase itself. Definitions used in this Test Guideline are given in Annex 1. INITIAL CONSIDERATIONS AND LIMITATIONS 3. Due to the limited number of chemicals tested and laboratories involved in the validation of this rather complex assay, especially inter-laboratory reproducibility is not documented with experimental data so far, it is anticipated that when a sufficient number of studies is available to ascertain the impact of this new study design, the Test guideline will be reviewed and if necessary revised in light of experience gained. The LAGDA is an important assay to address potential contributors to amphibian population declines by evaluating the effects from exposure to chemicals during the sensitive larval stage, where effects on survival and development, including normal development of reproductive organs, may adversely affect populations. 4. The test is designed to detect an apical effect(s) resulting from both endocrine and non-endocrine mechanisms, and includes diagnostic endpoints which are partly specific to key endocrine modalities. It should be noted that until the LAGDA was developed, no validated assay existed that served this function for amphibians. 5. Before beginning the assay, it is important to have information about the physicochemical properties of the test chemical, particularly to allow the production of stable chemical solutions. It is also necessary to have an adequately sensitive analytical method for verifying test chemical concentrations. Over a duration of approximate 16 weeks, the assay requires a total number of 480 animals, i.e., X. laevis embryos, (or 640 embryos, if a solvent control is used) to ensure sufficient power of the test for the evaluation of population-relevant endpoints such as growth, development and reproductive maturation. 6. Before use of the Test Guideline for regulatory testing of a mixture, it should be considered whether it will provide acceptable results for the intended regulatory purpose. Furthermore, this assay does not evaluate fecundity directly, so it may not be applicable for use at a more advanced stage than Level 4 of the OECD Conceptual Framework. SCIENTIFIC BASIS FOR THE TEST METHOD 7. Much of our current understanding of amphibian biology has been obtained using the laboratory model species X. laevis. This species can be routinely cultured in the laboratory, ovulation can be induced using human chorionic gonadotropin (hCG) and animal stocks are readily available from commercial breeders. 8. Like all vertebrates, reproduction in amphibians is under the control of the hypothalamic pituitary gonadal (HPG) axis (4). Oestrogens and androgens are mediators of this endocrine system, directing the development and physiology of sexually-dimorphic tissues. There are three distinct phases in the life cycle of amphibians when this axis is especially active: (1) gonadal differentiation during larval development, (2) development of secondary sex characteristics and gonadal maturation during the juvenile phase and (3) functional reproduction of adults. Each of these three developmental windows are likely susceptible to endocrine perturbation by certain chemicals such as estrogens and androgens, ultimately leading to a loss 2 © OECD, (2015) OECD/OCDE 241 of reproductive fitness by the organisms. 9. The gonads begin development at NF stage 43, when the bipotential genital ridge first develops. Differentiation of the gonads begins at NF stage 52 when primordial germ cells either migrate to medullary tissue (males) or remain in the cortical region (females) of the developing gonads (3). This process of sexual differentiation of the gonads was first reported to be susceptible to chemical alteration in Xenopus in the 1950's (5) (6). Exposure of tadpoles to estradiol during this period of gonad differentiation results in sex reversal of males that when raised to adulthood are fully functional females (7) (8). Functional sex reversal of females into males is also possible and has been reported following implantation of testis tissue in tadpoles (9). However, although exposure to an aromatase inhibitor also causes functional sex reversal in X. tropicalis (10), this has not been shown to occur in X. laevis. Historically, toxicant effects on gonadal differentiation have been assessed by histological examination of the gonads at metamorphosis and sex reversal could only be determined by analysis of sex ratios. Until recently, there had been no means to directly determine the genetic sex of Xenopus. However, recent establishment of sex linked markers in X. laevis make it possible to determine genetic sex and allows for the direct identification of sex reversed animals (11). 10. In males, juvenile development proceeds as blood levels of testosterone increase corresponding with the development of secondary sex characteristics as well as testis development. In females, estradiol is produced by the ovaries resulting in the appearance of vitellogenin (VTG) in the plasma, vitellogenic oocytes in the ovary and the development of oviducts (12). Oviducts are female secondary sex characteristics that function in oocyte maturation during reproduction. Jelly coats are applied to the outside of oocytes as they pass through the oviduct and collect in the ovisac, ready for fertilization. Oviduct development appears to be regulated by oestrogens as development correlates with blood estradiol levels in X. laevis (13) and X. tropicalis (12). The development of oviducts in males following exposure to polychlorinated biphenyl compounds (14) and 4-tert-octylphenol (15) has been reported. PRINCIPLE OF THE TEST 11. The test design entails exposing X. laevis embryos at NF stage 8-10 via the water route to four different concentrations of test chemical as well as control(s) until 10 weeks after the median time to NF stage 62 in the control with one interim sub-sample at NF stage 62. While it may also be possible to dose highly hydrophobic chemicals via the feed, there has been little experience using this exposure route in this assay to date. There are four replicates in each test concentration with eight replicates for each control used. Endpoints evaluated during the course of the exposure include those indicative of generalized toxicity (i.e., mortality, abnormal behavior and growth determinations (length and weight)), as well as endpoints designed to characterize specific endocrine toxicity modes of action targeting oestrogen-, androgen-, or thyroid-mediated physiological processes (i,e. thyroid histopathology, gonad and gonad duct histopathology, abnormal development, plasma vitellogenin (optional), and genotypic/phenotypic sex ratios) . TEST VALIDITY CRITERIA 12. The following criteria for test validity apply:  The dissolved oxygen concentration should be ≥ 40% of air saturation value throughout the test;  The water temperature should be in the range of 21 ± 1 °C and the inter-replicate and the inter- treatment differentials should not exceed 1.0 ºC;  pH of the test solution should be maintained between 6.5 and 8.5, and the inter-replicate and the inter-treatment differentials should not exceed 0.5; 3 © OECD, (2015) 241 OECD/OCDE  Evidence should be available to demonstrate that the concentrations of the test chemical in solution have been satisfactorily maintained within ± 20% of the mean measured values;  Mortality over the exposure period should be ≤ 20% in each replicate in the controls;  ≥ 70% viability in the spawn chosen to start the study;  The median time to NF stage 62 of the controls should be ≤ 45 days.  The mean weight of test organisms at NF stage 62 and at the termination of the assay in controls and solvent controls (if used) should reach 1.0 ± 0.2 and 11.5 ± 3 g, respectively. 13. While not a validity criterion, it is recommended that at least three treatment levels with three uncompromised replicates be available for analysis. Excessive mortality, which compromises a treatment, is defined as > 4 mortalities (> 20%) in 2 or more replicates that cannot be explained by technical error. At least three treatment levels without obvious overt toxicity should be available for analysis. Signs of overt toxicity may include, but are not limited to, floating on the surface, lying on the bottom of the tank, inverted or irregular swimming, lack of surfacing activity, and being nonresponsive to stimuli, morphological abnormalities (e.g., limb deformities), hemorrhagic lesions, and abdominal oedema. 14. In case a deviation from the test validity criteria is observed, the consequences should be considered in relation to the reliability of the test results, and these deviations and considerations should be included in the test report. DESCRIPTION OF THE METHODS Apparatus 15. Normal laboratory equipment and especially the following: (a) temperature controlling apparatus (e.g., heaters or coolers adjustable to 21 ± 1 ºC); (b) thermometer; (c) binocular dissection microscope and dissection tools; (d) digital camera with at least 4 megapixel resolution and micro function (if needed); (e) analytical balance capable of measuring to 0.001 mg or 1 µg; (f) dissolved oxygen meter and pH meter; (g) light intensity meter capable of measuring in lux units. Water Source and quality 16. Any dilution water that is locally available (e.g. spring water or charcoal-filtered tap water) and permits normal growth and development of X. laevis can be used, and evidence of normal growth in this water should be available. Because local water quality can differ substantially from one area to another, analysis of water quality should be undertaken, particularly if historical data on the utility of the water for raising amphibian larvae is not available. Measurements of heavy metals (e.g. Cu, Pb, Zn, Hg, Cd, Ni), major anions and cations (e.g. Ca, Mg, Na, K, Cl, SO ), pesticides, total organic carbon and suspended 4 solids should be made before testing begins and/or, for example, every six months where a dilution water is known to be relatively constant in quality. Some chemical characteristics of acceptable dilution water are listed in ANNEX 2. Iodide concentration in test water 4 © OECD, (2015) OECD/OCDE 241 17. In order for the thyroid gland to synthesize thyroid hormones to support normal metamorphosis, sufficient iodide should to be available to the larvae through a combination of aqueous and dietary sources. Currently, there are no empirically derived guidelines for minimum iodide concentrations in either food or water to ensure proper development. However, iodide availability may affect the responsiveness of the thyroid system to thyroid active agents and is known to modulate the basal activity of the thyroid gland which deserves attention when interpreting the results from thyroid histopathology. Based on previous work, successful performance of the assay has been demonstrated when dilution water iodide (I-) concentrations range between 0.5 and 10 μg/L. Ideally, the minimum iodide concentration in the dilution water throughout the test should be 0.5 μg/L (added as the sodium or potassium salt). If the test water is reconstituted from deionized water, iodine should be added at a minimum concentration of 0.5 μg/L. The measured iodide concentrations from the test water (i.e., dilution water) and the supplementation of the test water with iodine or other salts (if used) should be reported. Iodine content may also be measured in food(s) in addition to test water. Exposure system 18. The test was developed using a flow-through diluter system. The system components should have water-contact components of glass, stainless steel, and/or other chemically inert materials. Exposure tanks should be glass or stainless steel aquaria and tank usable volume should be between 4.0 and 10.0 L (minimum water depth of 10 to 15 cm). The system should be capable of supporting all exposure concentrations, a control, and a solvent control, if necessary, with four replicates per treatment and eight in the controls. The flow rate to each tank should be constant in consideration of both the maintenance of biological conditions and chemical exposure. It is recommended that flow rates should be appropriate (e.g., at least 5 tank turnovers per day) to avoid chemical concentration declines due to metabolism by both the test organisms and aquatic microorganisms present in the aquaria or abiotic routes of degradation (hydrolysis, photolysis) or dissipation (volatilization, sorption). The treatment tanks should be randomly assigned to a position in the exposure system to reduce potential positional effects, including slight variations in temperature, light intensity, etc. Further information on setting up flow-through exposure systems can be obtained from the ASTM Standard Guide for Conducting Acute Toxicity Tests on Test Materials with Fishes, Macroinvertebrates, and Amphibians (16). Chemical delivery: preparation of test solutions 19. To make test solutions in the exposure system, stock solution of the test chemical should be dosed into the exposure system by an appropriate pump or other apparatus. The flow rate of the stock solution should be calibrated in accordance with analytical confirmation of the test solutions before the initiation of exposure, and checked volumetrically periodically during the test. The test solution in each chamber should be renewed at a minimum of 5 volume renewals/day. 20. The method used to introduce the test chemical to the system can vary depending on its physicochemical properties. Therefore, prior to the test, baseline information about the chemical that is relevant to determining its testability should be obtained. Useful information about test chemical-specific properties include the structural formula, molecular weight, purity, stability in water and light, pKa and Kow, water solubility (preferably in the test medium) and vapour pressure as well as results of a test for ready biodegradability (OECD TG 301 (17) or TG 310 (18)). Solubility and vapour pressure can be used to calculate Henry's law constant, which will indicate whether losses due to evaporation of the test chemical may occur. Conduct of this test without the information listed above should be carefully considered as the study design will be dependent on the physicochemical properties of the test chemical and, without these data test results may be difficult to interpret or meaningless. A reliable analytical method for the quantification of the test chemical in the test solutions with known and reported accuracy and limit of detection should be available. Water soluble test chemicals can be dissolved in aliquots of dilution water at 5 © OECD, (2015) 241 OECD/OCDE a concentration which allows delivery at the target test concentration in a flow-through system. Chemicals which are liquid or solid at room temperature and moderately soluble in water may require liquid:liquid or liquid:solid (e.g., glass wool column) saturators (19). While it may also be possible to dose very hydrophobic test chemicals via the feed, there has been little experience using that exposure route in this assay. 21. Test solutions of the chosen concentrations are prepared by dilution of a stock solution. The stock solution should preferably be prepared by simply mixing or agitating the test chemical in dilution water by mechanical means (e.g. stirring and/or ultrasonication). Saturation columns/systems or passive dosing methods (20) can be used for achieving a suitably concentrated stock solution. The preference is to use a co-solvent-free test system; however, different test chemicals will possess varied physicochemical properties that will likely require different approaches for preparation of chemical exposure water. All efforts should be made to avoid solvents or carriers because: (1) certain solvents themselves may result in toxicity and/or undesirable or unexpected responses, (2) testing chemicals above their water solubility (as can frequently occur through the use of solvents) can result in inaccurate determinations of effective concentrations, (3) the use of solvents in longer-term tests can result in a significant degree of “biofilming” associated with microbial activity which may impact environmental conditions as well as the ability to maintain exposure concentrations and (4) the absence of historical data that demonstrate that the solvent does not influence the outcome of the study, use of solvents requires a solvent control treatment which has significant animal welfare implications as additional animals are required to conduct the test. For difficult to test chemicals, a solvent may be employed as a last resort, and the OECD Guidance Document on Aquatic Toxicity Testing of Difficult Substances and Mixtures should be consulted (21) to determine the best method. The choice of solvent will be determined by the chemical properties of the test chemical and the availability of historical control data on the solvent. In the absence of historical data, the suitability of a solvent should be determined prior to conducting the definitive study. In the event that use of a solvent is unavoidable, and microbial activity (biofilming) occurs, recommend recording/reporting of the biofilming per tank (at least weekly) throughout the test Ideally, the solvent concentration should be kept constant in the solvent control and all test treatments. If the concentration of solvent is not kept constant, the highest concentration of solvent in the test treatment should be used in the solvent control. In cases where a solvent carrier is used, maximum solvent concentrations should not exceed 100 μl/L or 100 mg/L (21), and it is recommended to keep solvent concentration as low as possible (e.g, < 20 μl/L) to avoid potential effects of the solvent on endpoints measured (22). Test animals Test species 22. The test species is X. laevis because this is: (1) routinely cultured in laboratories worldwide, (2) easily obtainable through commercial suppliers and (3) capable of having its genetic sex determined. Adult care and breeding 23. Appropriate care and breeding of X. laevis is described by a standardized guideline (23). Housing and care of X. laevis are also described by Read (24). To induce breeding, three to five pairs of adult females and males are injected intraperitoneally with human chorionic gonadotropin (hCG). Female and male specimens are injected with e.g., approximately 800-1000 IU and 500-800 IU, respectively, of hCG dissolved in 0.6-0.9% saline solution (or frog Ringer's solution, an isotonic saline for use with amphibians; www.hermes.mbl.edu/biologicalbulletin/compendium/comp-RGR.html ). Injection volumes should be about 10 µl/g body weight (~1000 µl). Afterwards, induced breeding pairs are held in large tanks, undisturbed and under static conditions to promote amplexus. The bottom of each breeding tank should have a false bottom of stainless steel mesh (e.g., 1.25 cm openings) which permits the eggs to fall to 6 © OECD, (2015) OECD/OCDE 241 the bottom of the tank. Frogs injected with hCG in the late afternoon will usually deposit most of their eggs by mid-morning of the next day. After a sufficient quantity of eggs is released and fertilized, adults should be removed from the breeding tanks. Eggs are then collected and jelly coats are removed by L- cysteine treatment (23). A 2% L-cysteine solution should be prepared and pH adjusted to 8.1 with 1 M NaOH. This 21 oC solution is added to a 500 mL Erlenmeyer flask containing the eggs from a single spawn and swirled gently for one to two minutes and then rinsed thoroughly 6-8 times with 21 °C culture water. The eggs are then transferred to a crystallizing dish and determined to be > 70% viable with minimal abnormalities in embryos exhibiting cell division. TEST DESIGN Test concentrations 24. It is recommended to use a minimum of four chemical concentrations and appropriate controls (including solvent controls, if necessary). Generally, a concentration separation (spacing factor) not exceeding 3.2 is recommended. 25. For the purposes of this test, results from existing amphibian studies should be used to the extent possible in determining the highest test concentration so as to avoid concentrations that are overtly toxic. Information from, for example, quantitative structure-activity relationships, read across and data from existing amphibian studies such as the Amphibian Metamorphosis Assay, TG231 (25) and the Frog Embryo Teratogenesis Assay - Xenopus (23) and/or fish tests such as OECD TG229, TG234 and TG236 (26) (27) (28) may contribute toward setting this concentration. Prior to running the LAGDA a range finding experiment may be conducted. It is recommended that the range-finding exposure is initiated within 24 hours of fertilization and continued for 7-14 days (or more, if needed), and the test concentrations are set such that the intervals between test concentrations are no greater than a factor of 10. The results of the range finding experiment should serve to set the highest test concentration in the LAGDA. Note that if a solvent has to be used, then the suitability of the solvent (i.e. whether it may have an impact on the outcome of the study) could be determined as part of the range finding study. Replicates within treatment groups and controls 26. A minimum of four replicate tanks per test concentration and a minimum of eight replicates for the controls (and solvent control, if needed) should be used (i.e., the number of replicates in the control and any solvent control should be twice as large as the number of replicates of each treatment group, to ensure appropriate statistical power). Each replicate should contain no more than 20 animals. While using the test method in this guideline, the minimum number of animals processed would be 15 (5 for NF stage 62 sub- sample and 10 juveniles). However, additional animals are added to each replicate to factor in the possibility for mortality while maintaining the critical number of 15. PROCEDURE Assay overview 27. The assay is initiated with newly spawned embryos (NF stage 8-10) and continues into juvenile development. Animals are examined daily for mortality and any sign of abnormal behavior. At NF stage 62, a larval sub-sample (up to 5 animals per replicate) is collected and various endpoints are examined (Table 1). After all animals have reached NF stage 66, i.e. completion of metamorphosis (or after 70 days from the assay initiation, whichever comes first), a cull is carried out at random (but without sub-sampling) to reduce the number of animals (10 per tank) (see paragraph 43), and the remaining animals continue exposure until 10 weeks after the median time to NF stage 62 in the control. At test termination (juvenile sampling) additional measurements are made (Table 1). 7 © OECD, (2015) 241 OECD/OCDE Exposure conditions 28. A complete summary of test parameters can be found in ANNEX 3. During the exposure period, dissolved oxygen, temperature, and pH of test solutions should be measured daily. Conductivity, alkalinity, and hardness are measured once a month. For the water temperature of test solutions, the inter-replicate and inter-treatment differentials (within one day) should not exceed 1.0 ºC. Also, for pH of test solutions, the inter-replicate and inter-treatment differentials should not exceed 0.5. 29. The exposure tanks may be siphoned on a daily basis to remove uneaten food and waste products, being careful to avoid cross-contamination of tanks. Care should be used to minimize stress and trauma to the animals, especially during movement, cleaning of aquaria, and manipulation. Stressful conditions/activities should be avoided such as loud and/or incessant noise, tapping on aquaria, vibrations in the tank. Duration of exposure to the test chemical 30. The exposure is initiated with newly spawned embryos (NF stage 8-10) and continued until ten weeks after the median time to NF stage 62 (≤ 45 days from the assay initiation) in control group. Generally, the duration of the LAGDA is 16 weeks (maximum 17 weeks). Initiation of assay 31. Parent animals used for the initiation of the assay should have previously been shown to produce offspring that can be genetically sexed (ANNEX 5). After spawning of adults, embryos are collected, cysteine-treated to remove the jelly coat and screened for viability (ASTM, 2004). Cysteine treatment allows the embryos to be handled during screening without sticking to surfaces. Screening takes place under a dissecting microscope using an appropriately sized eye dropper to remove non-viable embryos. It is preferred that a single spawn resulting in greater than 70% viability be used for the test. Embryos at NF stage 8-10 are randomly distributed into exposure treatment tanks containing an appropriate volume of dilution water until each tank contains 20 embryos. Embryos should be carefully handled during this transfer in order to minimize handling stress and to avoid any injury. At 96 hours post fertilization, the tadpoles should have moved up the water column and begun clinging to the sides of the tank. Feeding regime 32. Feed and feeding rate change during different life stages of X. laevis are a very important aspect of the LAGDA protocol. Excessive feeding during the larval phase typically results in increased incidences and severity of scoliosis (ANNEX 8) and should be avoided. Conversely, inadequate feeding during the larval phase results in highly variable developmental rates among controls potentially compromising statistical power or confounding test results. ANNEX 4 provides recommended larval and juvenile diet and feeding regimes for X. laevis in flow-through conditions, but alternatives are permissible providing the test organisms grow and develop satisfactorily. It is important to note that if endocrine- specific endpoints are being measured, feed should be free of endocrine-active substances such as soy meal. Larval feeding 33. The recommended larval diet consists of trout starter feeds, Spirulina algae discs and goldfish crisps (e.g., TetraFin® flakes, Tetra, Germany) blended together in culture (or dilution) water. This mixture is administered three times daily on weekdays and once daily on weekends. Tadpoles are also fed live brine shrimp, Artemia spp., 24-hour-old nauplii, twice daily on weekdays and once daily on the weekends starting on day 8 post-fertilization. The larval feeding, which should be consistent in each test vessel, 8 © OECD, (2015) OECD/OCDE 241 should allow appropriate growth and development for test animals in order to ensure reproducibility and transferability of the assay results: (1) the median time to NF stage 62 in controls should be ≤ 45 days and (2) a mean weight within 1.0 ± 0.2 g at NF stage 62 in controls is recommended. Juvenile feeding 34. Once metamorphosis is complete, the feeding regime consists of premium sinking frog food, e.g., Sinking Frog Food -3/32 (Xenopus Express, FL, USA) (ANNEX 4). For froglets (early juveniles), the pellets are briefly run in a coffee grinder, blender or crushed with a mortar and pestle in order to reduce their size. Once juveniles are large enough to consume full pellets, grinding or crushing is no longer necessary. The animals should be fed once per day. The juvenile feeding should allow appropriate growth and development of the organisms: a mean weight within 11.5 ± 3 g in control juveniles at the termination of the assay is recommended. Analytical chemistry 35. Prior to initiation of the assay, the stability of the test chemical (e.g., solubility, degradability, and volatility) and all analytical methods needed should be established e.g., using existing information or knowledge. When dosing via the dilution water, it is recommended that test solutions from each replicate tank concentration be analyzed prior to test initiation to verify system performance. During the exposure period, the concentrations of the test chemical are determined at appropriate intervals, preferably every week for at least one replicate in each treatment group, rotating between replicates of the same treatment group every week. It is recommended that results be based on measured concentrations. However, if concentration of the test chemical in solution has been satisfactorily maintained within ± 20% of the nominal concentration throughout the test, then the results can either be based on nominal or measured values. Also, the coefficient of variation (CV) of the measured test concentrations over the entire test period within a treatment should be maintained at 20% or less in each concentration. When the measured concentrations do not remain within 80-120% of the nominal concentration (for example, when testing highly biodegradable or adsorptive chemicals), the effect concentrations should be determined and expressed relative to the arithmetic mean concentration for flow-through tests. 36. The flow rates of dilution water and stock solution should be checked at appropriate intervals (e.g. three times a week) throughout the exposure duration. In the case of chemicals which cannot be detected at some or all of the nominal concentrations, (e.g., due to rapid degradation or adsorption in the test vessels, or by marked chemical accumulation in the bodies of exposed animals), it is recommended that the renewal rate of the test solution in each chamber be adapted to maintain test concentrations as constant as possible. Observations and endpoint measurements 37. The endpoints evaluated during the course of the exposure are those indicative of toxicity including mortality, abnormal behavior such as clinical signs of disease and/or general toxicities, and growth determinations (length and weight), as well as pathology endpoints which may respond to both general toxicity and endocrine modes of action targeting oestrogen-, androgen-, or thyroid-mediated pathways. In addition, plasma VTG concentration may be optionally measured at the termination of the assay. Measurement of VTG may be useful in understanding study results in the context of endocrine mechanisms for suspected EDCs. The endpoints and timing of measurements are summarized in Table 1. 9 © OECD, (2015) 241 OECD/OCDE Table 1. Endpoint overview of the LAGDA. Interim Sampling Test Termination Endpoints* Daily (Larval sampling) (Juvenile sampling) Mortality and abnormalities X Time to NF stage 62 X Histo(patho)logy (thyroid gland) X Morphometrics (growth in weight X X and length) Liver-somatic index (LSI) X Genetic/phenotypic sex ratios X Histopathology (gonads, X reproductive ducts, kidney and liver) Vitellogenin (VTG) (optional) X * All endpoints are analyzed statistically. Mortality and daily observations 38. All test tanks should be checked daily for dead animals and mortalities recorded for each tank. Dead animals should be removed from the test tank as soon as observed. The developmental stage of dead animals should be categorized as either pre-NF stage 58 (pre-forelimb emergence), NF stage 58-NF stage 62, NF stage 63-NF stage 66 (between NF stage 62 and complete tail absorption), or post-NF stage 66 (post-larval). Mortality rates exceeding 20% may indicate inappropriate test conditions or overtly toxic effects of the test chemical. The animals tend to be most sensitive to non-chemical induced mortality events during the first few days of development after the spawning event and during metamorphic climax. Such mortality could be apparent from the control data. 39. In addition, any observation of abnormal behavior, grossly visible malformations (e.g., scoliosis), or lesions should be recorded. Observations of scoliosis should be counted (incidence) and graded with respect to severity (e.g., not remarkable – NR, minimal – 1, moderate – 2, severe – 3; ANNEX 8). Efforts should be made to ensure that the prevalence of moderate and severe scoliosis is limited (e.g., below 10% in controls) throughout the study, although greater prevalence of control abnormalities would not necessarily be a reason for stopping the test. Normal behavior for larval animals is characterized by suspension in the water column with tail elevated above the head, regular rhythmic tail fin beating, periodic surfacing, operculating, and being responsive to stimuli. Abnormal behaviors would include, for example, floating on the surface, lying on the bottom of the tank, inverted or irregular swimming, lack of surfacing activity, and being nonresponsive to stimuli. For post-metamorphic animals, in addition to the above abnormal behaviors, gross differences in food consumption between treatments should be recorded. Gross malformations and lesions could include morphological abnormalities (e.g., limb deformities), hemorrhagic lesions, abdominal edema, and bacterial or fungal infections, to name a few. The occurrences of lesions on the head of juveniles, just posterior to the nostrils, may be indications of insufficient humidity levels. These determinations are qualitative and should be considered akin to clinical signs of disease/stress and made in comparison to control animals. If the rate of occurrence is greater in exposed tanks than in the controls, then these should be considered as evidence for overt toxicity. 10 © OECD, (2015)