Advertisement

The Human ADME Study

  • Andrew McEwenEmail author
Living reference work entry

Abstract

The human mass balance study is a pivotal study in the drug development process. Whilst a reasonable understanding of the absorption, distribution, metabolism and excretion (ADME) properties of the candidate drug will have been determined using pre-clinical models, the ultimate validation is provided following administration to human volunteers. The human ADME (hADME) study provides the link between pre-clinical safety studies and the clinical observations. Whilst described as a mass balance study, the key objective of the hADME study is the quantification, characterisation and identification of drug and drug metabolites present in systemic circulation. An assessment of the relative exposure between clinical subjects and the species used for pre-clinical safety studies enables a complete safety profile to be obtained.

Introduction

During the drug development process, considerable effort is expended in obtaining an understanding of the absorption, distribution, metabolism, and elimination (ADME) properties of drugs using, animal data, in vitro models and in silico models. While these models undergo constant revision and change to meet current regulatory requirements, the ultimate validation of these models comes with a pivotal study in the process – the human ADME study (hADME). In this study, all the predictions and measurements are tested using the target species, man. The study design usually covers a range of objectives but is essentially a metabolism study designed to investigate the routes and rates of excretion, pharmacokinetics in blood and plasma and confirm the identity of major circulating metabolites.

The regulatory question that the hADME study is intended to answer is often phrased as “Do you understand the metabolism of your compound in man?” but ultimately the question is better phrased “Do you understand the metabolism of your compound in man and how it relates to the metabolism in species selected for pre-clinical investigations?”

Before considering the modern human mass balance study, it is worth noting that the quantitative basis on which the study is reliant was established around 450 years ago by the work of Sanctorius. Sanctorius was a Venetian professor based at the University of Padua who is credited with numerous inventions including the thermometer. Over a period of 30 years he weighed himself, everything he ate or drank and also his urine and feces. He found that for every 3.6 kg of food he consumed only 1.4 kg of waste product were excreted. The difference was assigned to a process he described as “metabolism,” and a new science was established. The current human ADME study follows a similar experimental process; the drug administered is essentially weighed and administered to volunteers. The excreta is collected and weighed before the concentrations of drug are determined and a mass balance obtained.

The first recorded example of a human ADME study was provided Alexander Ure (1841). The study was inspired by the work of Woehler and Tiedemann (1824) that showed benzoic acid was converted to hippuric acid and eliminated in the urine following administration to dogs. Ure had a particular interest in the treatment of gout, and he proposed that the symptoms of gout may be relieved by reducing the synthesis of uric acid by administration of benzoic acid. He therefore conducted a human metabolism study and was successful in demonstrating the presence of hippuric acid in the urine collected following administration. The first successful human metabolism study may also be the first example of a failed therapeutic use as the administration of benzoic acid failed to reduce the concentrations of uric acid observed in the urine. The potential therapeutic use of benzoic acid in controlling gout was not discussed in the subsequent publication.

Strategies for quantifying drug metabolites has been debated widely, but despite the great advances in modern bioanalytical methods the use of radioisotopes is still an invaluable part of pharmaceutical development programs. The European Bioanalytical Forum presented a white paper outlining the issues associated with LC-MS/MS methods and providing some recommendations (Timmerman et al. 2011). The paper compared of responses obtained using LC-MS/MS with those obtained using conventional radiometric detection and highlighted the issues associated with differential ionization in the mass spectrometer. Examples were provided showing that in some cases there was a 30-fold difference in relative exposure between the two methods. The use of a radiolabeled test compound relies on no assumptions as to the identity of the drug-related compounds present in excreta or plasma and needs no internal standard and the response is structure independent allowing direct quantification.

Radioisotopes commonly used as tracers in metabolism experiments include [3H], [131I], [32P], and [35S] inserted as replacements for the hydrogen, iodine, phosphorus, or sulfur atoms commonly found in drugs; a summary of their radioactive properties is provided in Table 1. The isotope of choice, however, is [14C] due to long half-life, good detection efficiency, and the ability to place the radiolabel in a metabolically stable position within the molecule. Tritium is widely used in preclinical studies but suffers from low counting efficiency, short half-life, and the potential for isotope exchange or metabolic cleavage. Careful selection of the labelling site and adoption of targeted labelling strategies can help mitigate these liabilities (Lockley et al. 2012), but the use of tritium labels in hADME studies is still quite limited.
Table 1

Properties of radionuclides commonly used in hADME studies

Radionuclide

Emission

Half-life (years)

Specific activity

3H

Beta

12.3

28.8 Ci/mmol

14C

Beta

5730

62.4 mCi/mmol

32P

Beta

0.04

5118 mCi/mmol

35S

Beta

0.24

1494 mCi/mmol

133I

Beta

0.16

16382 Ci/mmol

Purpose and Rationale

Human metabolism studies such as those performed by Ure were initially focused on identification of transformation products in excreta (urine and feces), and it remained this way for over 100 years. Improvements in analytical methodologies meant that bioanalytical measurements could be performed using plasma to demonstrate that the preclinical species used in safety testing were exposed to drug material. Most bioanalytical measurements are currently performed using liquid chromatography-mass spectrometric techniques, and for the interested reader, the development of the discipline has been reviewed elsewhere (Hill 2009). As the sensitivity of modern instruments improved, the focus moved from identification of metabolites present in excreta to quantification and identification of the metabolites in circulation. The importance of plasma as the primary focus for metabolite identification studies was reinforced by the Metabolites in Safety Testing (MIST) guidelines introduced by the FDA in 2008 and the ICH in 2009 in response to a perceived industry need for clarification of the regulatory position.

Determination of the mass balance is included in the study design, but dependent on the PK/ADME properties of the test compound a full mass balance may not be obtained. Of greater significance is an assessment of the relative routes and rates of excretion. Similarly a determination of pharmacokinetic parameters for total radioactivity, parent compound, and metabolites in systemic circulation is a key activity included as part of the study plan. The key measurement here is the comparison of AUC’s obtained for total radioactivity and parent. This assesses the metabolite load and combined with radio-chromatographic analysis can validate the findings of the bioanalytical measurements performed in support of the clinical studies performed earlier in development.

The main purpose of the hADME study is therefore to provide a validation of the preclinical species used for safety testing, identify disproportionate or unique human metabolites, and “close the circle” between preclinical and clinical investigations.

Despite (or possibly because of) constant advances in analytical instrumentation with respect to sensitivity and resolution, it can be advantageous to position the timing of the hADME study as close as possible to the preclinical evaluations as possible. This enables preclinical and clinical samples to be analyzed on the same equipment at the same time. A direct comparison of retention times is possible and scouting for “unique” human metabolites in the toxicological species is possible.

The absence of direct regulatory guidance helped promote a strong debate as to which matrix was most important with regard to metabolite identification. In 2002 a multidisciplinary committee sponsored by the Pharmaceutical Research and Manufacturers of America published a report outlining best practice entitled “Drug Metabolites in Safety Testing” (Baillie et al. 2002). The primary trigger for further investigation was suggested as any component that accounted for 25% drug-related material in the systemic circulation. The report received a rapid response from the FDA who challenged the 25% trigger citing recommendations in the veterinary medicine guidelines where 10% of the total drug present was classified as a major metabolite. The draft regulatory “MIST” guidance document was published in 2005 and provided a trigger level of 10% drug-related material as the requirement for further investigation. The proposed guidance significantly increased the dialogue surrounding metabolite safety, and several further publications were generated reflecting opinion from within the industry (Smith and Obach 2005, 2006; Humphreys and Unger 2006; Prueksaritanont et al. 2006; Naito et al. 2007). Of particular interest was the proposition by Smith and Obach (2005) that absolute abundance and dose should be taken into consideration, noting that a metabolite representing 10% AUC obtained following a 1 g dose should be treated differently from a metabolite representing 10% AUC observed following a 1 mg dose due to the fact that the body burden would differ by a factor of 1000.

After considerable debate, the final FDA guidance was published in 2008 which contained a surprising change to the draft document where the trigger level for safety evaluation was set at 10% AUC of parent drug. For drugs that were extensively metabolized, this would mean that almost every component in circulation would require quantification and identification. The FDA position was undermined by the harmonized ICH document published in 2010 by the European Medicines Agency which re-iterated the 10% total drug-related material cut-off for metabolite investigations, and the FDA document was subsequently amended accordingly (FDA 2016). The revised guidance also makes reference to a document published by the EMEA in 2012 and notes this will represent the agencies current opinion on the subject. The definition of disproportionate metabolites was therefore harmonized, and the decision tree as now presented in the FDA guidance is reproduced in Fig. 1.
Fig. 1

FDA decision tree (Reproduced from US-FDA 2016)

Dosimetry

The hADME study is usually performed late in the drug development program, and there is generally a reasonable understanding of the chemical safety of the drug. The aim of the dosimetry study is therefore to assess the risk associated with administration of a radiolabeled test compound to human volunteers. The risk must be balanced with the objectives of the study, the aim being to administer sufficient radioactivity to achieve mass balance and fully characterize the proportions and identity of metabolites in systemic circulation.

The preclinical studies required to achieve this include a mass balance evaluation in the rat and a tissue distribution study to assess the relative contribution from each organ to the internal dose. The mass balance study provides information on relative rates and routes of elimination, and therefore, an assessment of the radioactive dose received by the gastrointestinal tract and bladder. The tissue distribution study can be performed using tissue excision and subsequent analysis of the tissues, but this method requires preselection of the tissues of interest prior to analysis and is heavily reliant on the skill of the technician performing the necropsy. The method of choice is now quantitative whole body autoradiography (Ullberg 1954), which relies on no preconceptions, provides information on the intraorgan distribution of radioactivity, and is nondestructive meaning that the tissue sections can be reanalyzed for radioactive content or taken for additional analyses such as tissue staining or MALDI experiments to determine the localization of both drug and metabolites (McEwen, Henson and Wood 2014).

Most tissue distribution studies are performed using the same strain of albino rats as chosen for the toxicology evaluations, thus enabling findings to be correlated with the presence of drug related material. The investigations also tend to run for up to 7 days reflecting the timeline for the mass balance studies. This design has two drawbacks when estimating the dose administered to man. There are no pigmented tissues in which to assess the binding of the drug to melanin and there may still be tissues with radioactivity present at the last timepoint. From a chemical safety standpoint, there is no indication that melanin binding is in any way associated with toxicity (LeBlanc et al. 1998), but binding of a radioactive drug can provide a significant internal dose to the eye. While the eye is not one of the mandatory tissues for dosimetric calculation as listed by the ICRP, it is therefore prudent to include it in the risk assessment.

The second deficiency in the standard QWBA experiment when calculating the internal dose in man is that the presence of radioactivity in tissues at the final sampling point triggers the use of standard half-life values in the dosimetry calculations. These are generally taken as 100 h in the USA and a more conservative 100 days in the EU. Extending the experiment to 21, 28, or 35 days can demonstrate elimination of radioactivity from the tissues or alternatively provide sufficient data with which to calculate a biological half-life.

A description of the dosimetry calculation has been presented in depth in an earlier chapter in this series (Kuerzel et al. 2011) and so will not be discussed further here. The aim is to assess the risk to human volunteers following administration of a radioactive test substance, and these are related to guidance provided by the WHO (1977) and subsequent modification by the ICRP (1992).

The guidance takes into account the total risk associated with exposure, namely the probability of a fatal cancer, nonweighted probability of a nonfatal cancer, and probability of successive generations suffering serious hereditary disease as a result of administration of the radiolabeled test item. The risk categories are reproduced in Table 2.
Table 2

Risk categories based on radioactive dose (ICRP)

Risk level

Risk category

Corresponding effective dose range in adults (mSv)

Level of societal benefit

Trivial

Category I

<0.1

Minor

Minor to Intermediate

Category IIa

0.1–1.0

Intermediate to moderate

Category IIb

1.0–10

Moderate

Category III

>10

Substantial

The risks run from category I where the risk is one in a million to category III where the risk is one in a thousand. Exposure of volunteers to an effective dose between 0.1 and 1.0 mSv is generally considered acceptable for biomedical investigations.

When preparing the preclinical package, it is worth considering the clinical study when designing the mass balance and tissue distribution studies. The routes and rates of excretion can be taken from albino animals, and if a full balance is not obtained, then a ratio of urinary to fecal excretion can be employed. The data should be generated using the same sex animals and dose route as intended in the hADME study. A list of the tissues specified for dosimetric evaluation is provided in Table 3. It should be noted that the remainder is then provided as a separate list. Tissue distribution studies are generally performed using albino animals, the same strain used in toxicological safety studies, but for assessment of the risk to man it is advisable to use a pigmented strain for measurement of melanin binding. The additional tissues are provided in Table 4. It can be seen that some of these tissues are quite small and the use of QWBA overcomes the potential contamination issues associated with traditional necropsy methods.
Table 3

Tissues for dosimetry

Tissue or organ

Weighting factor WR

Gonads

0.20

Bone Marrow (red)

0.12

Colon

0.12

Lung

O.12

Stomach

0.12

Bladder

0.05

Breast

0.05

Liver

0.05

Oesophagus

0.05

Thyroid

0.05

Skin

0.01

Bone surface

0.01

Remainder

0.05

Table 4

Additional tissues

Adrenals

Myocardium

Adipose tissue

Pancreas

Blood

Pituitary gland

Brain

Prostate

Eye (uveal tract)

Small intestine

Upper large intestine

Spleen

Kidney

Thymus

Muscle

Uterus

Radiolabeled Test Compound

A key factor in the successful completion of a hADME study is accurate quantification of the dose material. In 2001 the EU passed laws relating to the implementation of good clinical practice in the conduct of clinical trials on medicinal products for human use and the application of good manufacturing practice (GMP) was extended to investigational products. The test compound should have a chemical and radiochemical purity greater than 98%, a well-characterized specific activity, and a stability trial should be performed to assure product integrity between synthesis and dosing.

Clinical specific activity (μCi/mg) = Amount of radioactivity (μCi)/Amount of drug (mg)

The drug material administered should be homogeneous and from a single batch. Experiments, for example, where cold material is administered in one capsule and the radioactive dose is administered in a separate capsule can fail due to differential absorption of different crystalline forms of the test material resulting in an absorbed dose with unknown specific activity.

The synthesis of radiolabeled compounds for use in clinical studies is covered in a separate chapter (Atztrodt) and will not be discussed further here. For investigational products administered intravenously, further work may be required prior to administration to assess binding to dosing cannula.

Clinical Study Design

Study Title

A general title for the human ADME study can be expressed as follows:

A phase 1, open label study investigating the absorption, metabolism, and elimination (AME) of 14C-ABC123 following a single oral dose to healthy male subjects.

It should be noted that “distribution” is missing in this title reflecting a difference of opinion between investigators as to whether distribution can be determined as part of the radiolabeled human mass balance study. The study is also open meaning that the identity of test compound administered is known to the investigators performing the analysis.

Objectives

When preparing a study protocol for the human ADME study, the objectives are generally categorized as primary and secondary. As an example the objectives could be listed as follows:

The primary objective of the study is to determine the pharmacokinetics of total radioactivity in plasma, whole blood, and red blood cells along with the pharmacokinetics of ABC123 following a single oral dose of 10 mg ABC123

The secondary objectives are:
  • To determine the relative routes and rates of excretion and obtain a mass balance by measuring the urinary and fecal excretion of radioactivity.

  • To estimate the protein binding of ABC123-related radioactivity in human plasma.

  • To obtain samples for quantification and identification of components present in systemic circulation.

  • To determine the safety and tolerability of a single dose of 10 mg 14C labelled ABC123 in healthy volunteers.

The primary objective is therefore to determine the systemic exposure of radioactivity following administration and relate that to the exposure to ABC123. The difference between total radioactivity and parent compound in systemic exposure is described as the metabolite load. If the exposure to total radioactivity and parent compound are the same (i.e., no metabolism), then there is no further work required. If the difference between total radioactivity and parent compound exposure is large, then further work is required to determine the proportions and identity of the radioactive components in addition to parent that are present. These investigations can sometimes be complex and time-consuming. For that reason, the metabolite identification activities are usually conducted and reported separately from the clinical investigations.

Study Design

The study design is generally presented as a single center, open label, single administration of 14C-ABC123 as an oral solution administered to fasted healthy male volunteers. As the study requires the use of a radioactive compound, female volunteers are generally not used in hADME studies. If the intended therapeutic use is for female conditions only, then the use of female volunteers would be required. In this case, the volunteers would be postmenopausal women.

Human Volunteers

The study would be conducted in healthy male subjects aged 30–55 years. Six subjects would be enrolled in the study with the expectation that at least four subjects would complete the study. Replacements are generally not required unless greater than two volunteers drop out. The process includes enrolment of eight subjects to present at clinic for screening on Day 1. Two subjects are eventually sent home once six subjects have successfully been dosed.

Dose Administration

A single oral dose of 10 mg 14C-labelled ABC123 will be given on the morning of Day 1 after an overnight fast. The radioactivity administered will correspond to an effective radiation dose of 0.1 to 1.0 mSv. The dose has been selected based on preclinical and clinical data and is considered safe and suitable for characterization of the pharmacokinetic properties of the drug.

The effective radiation dose is within the limits specified under the ICRP and WHO guidelines for administration of radioactivity to human volunteers.

Selection Criteria

The study plan generally lists specific criteria by which subjects can be enrolled, and these are split into inclusion and exclusion criteria.

Inclusion

Subjects who meet the following criteria can be admitted on the study:
  1. 1.

    The subject is able to read and understand the Volunteer Information sheet.

     
  2. 2.

    The subject has signed the study specific Informed Consent Form.

     
  3. 3.

    The subject is male.

     
  4. 4.

    The subject is between 40 and 55 years of age.

     
  5. 5.

    The subject has a minimum weight of 60 kg and a BMI between 19 and 29 kg/m2. If we assume typical sample sizes of 1 g and 0.25 g for determination of radioactive concentrations in urine and feces, respectively, then the detection limits (expressed as % administered dose) are defined by the following graphs.

     
  6. 6.

    The subject has a resting pulse in the normal range of 51–100 bpm.

     
  7. 7.

    The subject has a resting systolic blood pressure within the normal range of 91–179 mmHg (supine) and a resting diastolic blood pressure of 51–100 mmHg. The subject should also have an orthostatic blood pressure less than 20 mmHg.

     
  8. 8.

    The subject is considered in good health based on the results of a prestudy physical investigation, medical history, vital signs, an electrocardiogram and laboratory investigations including blood biochemistry, hematology, serology, and urinalysis within acceptable range.

     
  9. 9.

    For some compounds, the subject must agree to the use of a suitable barrier method of contraception throughout the study and for 3 months following study completion.

     

Exclusion

Subjects who meet one or more of the following criteria cannot be admitted on the study:
  1. 1.

    Use of concomitant medicine, specifically any prescribed medicine within 2 weeks of the study, or over the counter (OTC) medication within 1 week prior to dosing. Patients who have taken nonprescribed or topical medication may still be allowed on the study if in the opinion of the investigator, the medication will have no effect on the outcome of the study.

     
  2. 2.

    Subjects who at the screening visit fail a urinary drugs of abuse test or alcohol breath test.

     
  3. 3.

    Significant history of drug or alcohol abuse within the last 6 months. For assessment of alcohol consumption, one unit is described as ½ a pint of beer/lager, 1 glass of wine, or 1 shot of spirits.

     
  4. 4.

    The subject has participated in a clinical trial within 3 months prior to dosing.

     
  5. 5.

    The subject has a known sensitivity to ABC123.

     
  6. 6.

    The subject has a history of severe drug allergy or hypersensitivity.

     
  7. 7.

    The subject has a serious illness, such as renal or liver disfunction, or a cardiovascular, pulmonary, gastrointestinal, endocrine, neurological, infectious, neoplastic, or metabolic disorder.

     
  8. 8.

    A history of seizures.

     
  9. 9.

    The subject has been classified as either a “poor” or “fast” metabolizer for CYPD6.

     
  10. 10.

    The subject has tested positive for HIV.

     
  11. 11.

    Blood donation within 3 months prior to dosing.

     
  12. 12.

    A history of smoking or use of nicotine substitution therapy (patches, gum, or inhalers).

     
  13. 13.

    Positive test to a screen for drugs of abuse (amphetamines, barbiturates, benzodiazepines, cannabinoids, cocaine, methadone, or opiates).

     
  14. 14.

    The presence of clinically relevant cardiovascular disease including signs of arrhythmia/tachycardia of QT prolongation.

     
  15. 15.

    Receipt of an X-ray (other than dental X-rays) or radiolabeled material within 12 months preceding the study.

     
  16. 16.

    Occupational exposure to radioactive substances.

     
  17. 17.

    An irregular defection pattern <1, >3 defecation per day.

     
  18. 18.

    An assessment that the subject may be unwilling to comply with the clinical study protocol.

     

Withdrawals and Replacements

Subjects may be withdrawn from a study for a variety of reasons, these include:
  1. 1.

    Withdrawal of consent.

     
  2. 2.

    The investigator decides that the subject should be withdrawn for safety reasons.

     
  3. 3.

    A serious adverse event (SAE) occurs.

     
  4. 4.

    The subject is lost on follow-up.

     

The date and reason for withdrawal should be noted and, where possible, subjects that withdraw should be seen for a final evaluation and completion of the records.

Subjects that withdraw prior to dose administration can usually be replaced, whereas subjects who withdraw following dose administration are generally not replaced.

Restrictions

The study protocol generally lists a set of restrictions that must be followed by the volunteers during the conduct of the study. Standard exclusions include:

Alcohol: Subsects are usually asked to abstain from alcohol intake from 48 h prior to dosing until after the follow-up visit.

Caffeine- and/or xanthine -containing products: Subjects should refrain from consuming caffeine and/or xanthine from 48 h prior to dosing until they have completed the study. Typical xanthine-containing products include tea, coffee, cola, chocolate, and chewing gum.

Contraception: As the study involves the use of 14C-labelled material, the subjects will be asked to use a suitable method of contraception (such as a condom) throughout the study and for 3 months following study completion.

Exercise: Subjects are asked to refrain from strenuous exercise from 72 h prior to dosing until after the follow-up visit.

Grapefruit, grapefruit-containing products, marmalade, and Seville oranges: Subjects are asked to refrain from consuming grapefruit, grapefruit-containing products, marmalade, and Seville oranges from 48 h prior to dosing until after the follow-up visit.

Meals: Subjects are provided with a standard diet during their time in the clinic. Subjects usually fast from 10 h prior to dosing and should refrain from taking water from 2 h prior to dosing until 2 h postdose. After dosing, subjects may resume their usual rate of fluid consumption but may be limited to 2 l of water per day.

Smoking: Subjects are not permitted to smoke or use nicotine containing products throughout the study.

Discharge Procedures

The subjects will remain in the unit for a minimum of 120 h postdose (Day 5). After this period, collection of urine and feces will be discontinued and subjects will be released if the following criteria are met:
  • The combined cumulative excretion of radioactivity in urine and feces exceeds 90% of the administered dose.

  • The total radioactivity in two consecutive 24 h collections is below 1% of the administered dose.

If by Day 5 the release criteria are not met, the subjects can be kept in the unit until the release criteria are met or until Day 14 whichever comes first.

Duration of the Study

Subjects will remain in the clinic until Day 5 or until the release criteria are met. Further samples may be taken up to a maximum of 15 days.

Some study designs treat the subjects as a group and base release criteria upon the mean excretion from the group. This can lead to extended stays for volunteers if one of the groups has difficulty defecating or fails to comply with the collection procedures. An alternative strategy is to base the release criteria on individual excretion. This promotes compliance with the collection process as the subject can be released based on their personal criteria.

Pharmacokinetic Assessments

Blood Samples for Parent Drug Analysis and Total Radioactivity Determination

A series of samples will be taken for determination of total radioactivity and parent drug concentration throughout the study period. The exact sampling times will be recorded in the data.

Blood samples for determination of ABC123 will be drawn at predose and 1, 2, 4, 6, 7, 8, 9, 10, 12, 15, 24, 36, 48, 72, 96, and 120 h postdose.

Blood samples for measurement of radioactivity in plasma, whole blood, and red blood cells will be drawn at predose and 1, 2, 4, 6, 7, 8, 9, 10, 12, 15, 24, 36, 48, 72, 96, and 120 h postdose.

Overall 17 × 2 mL blood samples will be taken for determination of ABC123.

In addition, 17 × 6 mL blood samples will be taken for measurement of total radioactivity in plasma, whole blood, and red blood cells.

Subjects will be cannulated up to 24 h postdose to avoid repeated venipuncture. For each sample taken via cannula, the first 0.5 mL blood is discarded and following blood collection the cannula is flushed with saline.

Blood Samples for Protein Binding and Metabolite Investigations

A series of samples will be taken for determination of plasma protein binding and metabolite investigations throughout the study period.

Blood samples for determination of plasma protein binding will be drawn at predose and 4, 8, 12, and 24 h postdose.

Blood samples for metabolite investigations will be drawn at predose and 4, 8, 12, 24, 48, and 72 h postdose.

Overall 5 × 7 mL blood samples will be taken for determination of plasma protein binding.

In addition 7 × 10 mL blood samples will be taken for metabolite investigations.

Total Volume of Blood Drawn

Based on the sampling regime outlined above, the estimated volume of blood drawn from each subject would be around 280 mL over the course of the study.

The distribution of samples for each subject is as follows:
  • 8.1 mL for hematology (3 × 2.7 mL)

  • 9.0 mL for blood biochemistry and serology (1 × 9 mL)

  • 14 mL for blood biochemistry (2 × 7 mL)

  • 34 mL for drug assay (17 × 2 mL)

  • 35 mL for determination of plasma protein binding

  • 70 mL for metabolite investigations (7 × 10 mL)

  • 6 mL for discard during sampling via cannula.

Urine Sampling for Total Radioactivity Determination and Metabolite Investigations

Urine samples will be collected at the following intervals: −24–0, 0–6, 6–12, 12–24, 24–48, 48–72, 72–96, 96–120, 120–144, and 144–168 h postdose. If required further samples can be taken covering the intervals 168–192, 192–216, and 216–240 h postdose. The actual start and end times for each collection will be recorded in the study data along with the weight.

Feces Sampling for Total Radioactivity Determination and Metabolite Investigations

Feces samples will be collected at the following intervals: −24–0, 0–24, 24–48, 48–72, 72–96, 96–120, 120–144, and 144–168 h postdose. If required further samples can be taken covering the intervals 168–192, 192–216, and 216–240 h postdose. The actual start and end times for each collection will be recorded in the study data along with the weight.

Analytical Instrumentation

Radioactivity Determination

Radioactivity determinations are performed using a liquid scintillation counter (e.g., TriCarb model 2300TR Perkin Elmer). For plasma and urine analysis, duplicate weighed aliquots are mixed with scintillation cocktail (Ultima Gold XR) and analyzed directly by liquid scintillation counting. Duplicate aliquots of fecal homogenates and blood samples are weighed and combusted using an automatic sample oxidizer (Model 307, Perkin Elmer). The resultant [14CO2] is trapped in CarboSorb (Perkin Elmer) in combination with PermaFluor and radioactive content determined using liquid scintillation counting. Detected counts per minute (cpm) are converted to disintegrations per minute (dpm) using quench correction. The quench curves were prepared using standards purchased from Perkin Elmer Life and Analytical Sciences and are prepared from stock solutions that are calibrated against National Institute of Standards and Technology (NIST) Reference Materials. The validity of the curves is checked regularly throughout the study.

Measurement of ABC123 in Plasma

Plasma concentrations of ABC123 will be determined following protein precipitation using a liquid chromatography with tandem mass spectrometry (LC/MS/MS) assay validated for concentrations between 5.00 and 5000 ng/mL, with quality control (QC) samples prepared at 12.5, 750, and 3500 ng/mL. ABC123 and its deuterated analog D6-ABC123 (IS) will be extracted from plasma and concentrations of ABC123 determined.

Control blank human plasma containing K2EDTA as an anticoagulant will be obtained and stored at nominally −20 °C when not in use. Stability of ABC123 in human plasma will be determined.

Chromatographic Analysis

Sample Preparation

Urine

Equal proportions (between 1 and 5%) of each total sample weight will be combined to produce two time points per subject (0–4 and 4–24 h). All subsamples will be stored at around −70 °C. Aliquots of each subsample (100 μL) will be transferred into an Eppendorf tube and deionized water (400 μL) will be added before vortex mixing for ca. 5 s. The Eppendorf tube will then be centrifuged at 12,000 rpm at 4 °C for 5 min to sediment out any particulate matter before supernatant (140 μL) is transferred to a glass autosampler vial; the sample is ready for analysis.

Plasma

Plasma subsamples obtained in the study will be pooled per subject (1–24 h). All subsamples will be stored ca −70 °C prior to analysis.

Samples for extraction (2 mL) will be diluted with an equal volume of 10 mM ammonium acetate (aq). The diluted sample will then be subjected to the following SPE procedure using a Phenomenex Strata C18-E SPE cartridge (200 mg/3 mL):-
  • Prime: methanol (2 mL) followed by 10 mM ammonium acetate (aq) (2 mL)

  • Load: diluted sample (2 × 2 mL aliquots)

  • Wash: 10% methanol in 10 mM ammonium acetate (aq) (2 × 1 mL aliquots)

  • Elute: methanol:water (9:1 v/v) (2 × 2 mL aliquots)

The volume of the eluate will be reduced to ca. 400 μL under a gentle stream of nitrogen. The proportion of radioactivity extracted will be measured.

An aliquot of each extract (230 μL) will then be transferred into a glass autosampler vial. The vial will then be centrifuged at 12,000 rpm at 4 °C for 5 min to sediment out any particulate matter before being capped for analysis.

Feces

Feces subsamples obtained will be analyzed at two time points per subject, generally 24–48 h and 48–96 h: All subsamples will be stored at ca −70 °C prior to analysis.

Samples for extraction (ca. 5 g) will be combined with 10 ml water pH 4.0/Acetonitrile (20/80 v/v). Samples will be vortex mixed for ca. 1 min and then sonicated for ca. 5 min. The extraction step will be repeated a further two times. Samples will then be centrifuged for 10 min at ca. 8500 rpm and the supernatant decanted into a clean tube. Samples will be extracted up to three times and the extraction efficiency determined.

An aliquot of each sample (200 μL) will be transferred into an Eppendorf tube and deionized water (600 μL) added before vortex mixing for ca. 5 s. The Eppendorf tube will then be centrifuged at 12,000 rpm at 4 °C for 5 min to sediment out any particulate matter before the supernatant (230 μL) is transferred to a glass autosampler vial; the sample is ready for analysis.

Chromatography

Analysis of [14C]-ABC123 and its metabolites in the pooled urine, feces, and plasma samples will be performed using a gradient elution method . The equipment will consist of a Surveyor liquid chromatography system (Thermo Fisher Scientific, UK) equipped with an autosampler and a variable wavelength detector operating at 254°nm. Separations will be performed using reverse-phase chromatography on a YMC-Pack ODS-AQ 150 × 4.6 mm, 3μm, 12nm column (Crawford Scientific, UK) fitted with a C18 guard column (Phenomenex, UK). [14C]-ABC123 and its metabolites will be eluted from the column using a gradient based on two solvent mixtures:

Mobile phase A:

10 mM Ammonium acetate (pH 4.0)

Mobile phase B:

100 mM Ammonium acetate:acetonitrile (1:9 v/v)

The gradient will be established from 90% A at zerotime and held for 2 min, then the proportion of A decreased to 60% in a linear manner up to 60 min. The column will then be washed with 100% B for 5 min, returned to the initial conditions (60% A) and conditioned for 4 min prior to further injections. The flow rate will be 1.0 mL/min with a 4:1 split between the fraction collector and the mass spectrometer giving an approximate 200 μL/min flow to the mass spectrometer. Fractions will be collected every 13 s into 96-well Scintiplates (Perkin Elmer) for the duration of the analytical run using a modified CTC HTX PAL fraction collector (Presearch, UK). The tray temperature for holding samples will be set at 10 °C, while the column is held at 30 °C. Injection volumes will typically be 50–200 μL dependent upon sample.

Metabolites will be characterized using a Finnigan TSQ Quantum Ultra AM mass spectrometer (Thermo Scientific, UK), equipped with an electrospray source. Identification was performed using an LTQ Orbitrap XL mass spectrometer (Thermo Scientific, UK) and confirmed using accurate mass. The mass spectrometers will operated in the positive and/or negative ionization mode as appropriate. The instrument settings and potentials will adjusted as necessary to provide optimal data.

Review of the Method

Mass Balance

Following the publication of the regulatory guidelines in 2006, the recoveries obtained in clinical studies performed pre- and postguidance were reviewed to see if there was any effect on the recoveries obtained (McEwen et al. 2012). Urine and sample weights obtained from the volunteers were collated and the following mean values obtained:

Daily urine excreted (mL)

2283 ± 413

Daily feces excreted (g)

182 ± 19.2

If we assume typical sample sizes of 1 g and 0.25 g for determination of radioactive concentrations in urine and feces, respectively, then the detection limits (expressed as % administered dose) are defined by the following graphs (Fig. 2).
Fig. 2

Detection limits in urine and faeces based on dose administered

Radioactive doses administered in the studies were in the range 0.48–11.1 MBq (13–300 μCi). Even after administration of the lowest dose, the detection limits in urine and feces enabled determination of less than 0.1% dose per subject per day, thus easily enabling the relative routes of elimination to be adequately characterized. Beyond the 100 μCi dose (a value commonly employed in hADME studies), the increase in sensitivity provides no further advantage with regards to obtaining a mass balance and cannot be justified on this basis.

The recoveries obtained pre- and postguideline are provided in Table 5. In all cases, radioanalysis was performed in real time with results available within 8 h of sampling. The mean recovery obtained preguideline was 90.9 ± 8.0%, while the recovery postguideline was 91.8 ± 8.0%; therefore, implementation of the guidance had no notable effect on the recoveries obtained. It is notable however that the proportion of studies with >95% recovery decreased from 40% to 30% after 2008, but this is attributable to the fact that the release criteria were changed. The original release criteria were set at >95% recovery or <1% in 2 consecutive 24 h collections of excreta. This was amended to allow release at 90% recovery or 1% in two consecutive 24-h collections which explains the apparent fall in recoveries greater than 95%.
Table 5

Recoveries pre- and post-MIST

Recoveries

Pre-Mist 2002–2008

Post-MIST 2008–2012

Overall

Mean 14C Recovery (%)

90.9 ± 8.0

91.8 ± 8.0

91.3 ± 7.9

Proportion of studies with

>95% recovery

40.0

30.5

35.7

90–95% recovery

26.7

38.5

32.1

80–90% recovery

26.7

23.1

25.0

<80% recovery

6.7

7.7

7.1

Recoveries obtained pre- and post-MIST are shown in Fig. 3. The recoveries were initially presented based on total recovery but were then re-evaluated based on the route of excretion. Three categories were selected, >75% radioactivity excreted in urine, >75% radioactivity excreted in feces, and “mixed” where the excretion was not primarily urinary or fecal. There was no clear relation between route of excretion and the recovery obtained.
Fig. 3

Recoveries pre- and post- MIST

Recoveries greater than 80% were achieved in 93.3% studies pre-MIST and 92.3% studies post-MIST. Overall 94.1% of studies performed over the review period achieved recoveries greater than 80%; 90% recovery was achieved in 66.7% studies pre-MIST and 69.2% studies post-MIST.

Key factors for obtaining a good mass balance were identified as:
  • Accurate determination of the dose administered

  • Complete collection of the urine/feces samples

  • The PK/ADME properties of the test compound

Metabolite Quantification and Identification

The hADME study provides valuable samples of urine, feces, and plasma which can provide useful information on the metabolic fate of the test compound in man. While the samples are initially assayed for radioactive content, with the aim of obtaining a mass balance, the main utility of the samples is in the quantification and identification of drug metabolites.

Due to the low quantities of plasma that can be taken during the hADME study, initial method development is usually performed using samples of urine and feces. This enables establishment of optimal chromatographic conditions prior to analyzing the precious plasma extracts. Excreta samples containing notable quantities of radioactivity (usually >90% excreted radioactivity) are selected for analysis and can be pooled by timepoint and subject as appropriate. The standard analytical method for metabolite investigations is liquid chromatography with radioactivity detection. The use of radioactivity makes no assumptions as to the fate of the drug entity and as the response is independent of structure relative proportions of all components in the sample is achieved immediately.

The key objective of the hADME study is quantification and identification of the radioactive components in systemic circulation in order to meet the regulatory requirements of the ICH and FDA. As well as the limited amount of sample available, these investigations are also challenging due to fact that concentrations of radioactivity are generally much lower in plasma than in any other matrix. During the preclinical studies performed early in the drug development process, relatively high doses of radioactivity can be administered (50–100 μCi/kg) and the sample volumes are much lower. In the hADME study, the amount of radioactivity that can be administered is determined by the dosimetry assessment and for [14C] is usually around 50–100 μCi. Where drugs exhibit poor bioavailability, extensive metabolism, or high volumes of distribution, this can result in extremely low concentrations of drug and thus radioactivity in circulation. The analytical challenge is compounded by the definition what constitutes a major radioactive component requiring identification. The initial regulatory guidance provided by the FDA indicated that a major radioactive component would be one that accounts for >10% AUC of parent AUC, the subsequent guideline produced by the ICH indicated that a major radioactive component was one that accounted for >10% total radioactivity AUC. For some time, this led to confusion and uncertainty with an unstated belief that the ICH guideline would be the standard that took precedence. The issue was resolved by the publication of an updated guidance from the FDA (FDA 2016) in which the definition of a major radioactive component was confirmed as one that accounted for >10% total radioactivity AUC.

In contrast to the determination of the mass balance of the test compound where measurements are expressed as % dose administered, the detection limits in plasma are dependent upon the specific activity of the test material. A standard bioanalytical assay developed for plasma will generally rely upon the amount of chemical present in the matrix, whereas for radioactive studies the detection limit increases as the chemical dose increases. This situation is shown graphically for different radioactive doses in Fig. 4. Taking 100 mg as the proposed human dose, the detection limit decreases as the dose given to the volunteers is increased. For a standard radioactive dose of 100 μCi, the detection limit increases as the proposed chemical dose increases. This should be borne in mind when reviewing the dosimetry data – will the proposed radioactive dose meet the objectives of the study – and can be used as a justification when moving from ICRP category I dose to a category IIa dose. It is also worth noting that low radioactive doses do not necessarily result in low specific activity material. Doses prepared for studies supported by accelerator mass spectrometry (AMS, discussed later in the chapter) may contain a low radioactive dose but also tend to be administered as a low chemical dose, thus resulting in a high specific activity. This has knock on effects on assessing the potential stability of the dose material.
Fig. 4

Detection limits in plasma based on dose administered

Plasma samples can be prepared for quantification in a number of ways but are generally either across subjects at specific timepoints or more commonly by preparation of AUC pools using the methods published by Hamilton (Hamilton et al. 1981) or Hop (Hop et al. 1998). The AUC method provides information on the relative exposure of each metabolite compared to circulating total radioactivity, but can result in dilution of the radioactive response. In general, a mixed approach would be employed with an AUC pool supplemented by a couple of timepoint-specific analyses. AUC pools can be generated in a time- and volume-dependent manner across subject.

The gold standard approach to metabolite investigations using plasma samples from hADME studies has been the use of high performance liquid chromatography (HPLC) combined with radiodetection. In combination with modern high resolution mass spectrometers, the eluent from the column can be split (usually 10:90 or 20:80) to enable simultaneous quantification and identification of drug metabolites.

A major limitation to the use of radioflow detection methods is the poor sensitivity obtained. There are two main types of radioflow cell: those using a solid scintillant and those using a liquid flow cell where eluent is mixed with scintillator in the detector and counted. Of the two cell types, the flow through cell provides the best performance, but the key limiting factor for both detectors is the short residence time in the cell. Components in samples containing low concentrations of radioactivity can pass through the detector without providing sufficient counts for detection.

Recent detectors such as the BetaRam5 detector (LabLogic) use “active counting” (ACMTM) to provide improvements in the signal observed. An evaluation of the technique was reported by Attwood et al. (2010) and indicated consistent retention times and compatibility with UPLC. The use of ACMTM avoids the requirement for the sample to be fraction collected and counted off-line thereby eliminating the possible loss of volatile metabolites during sample processing.

Given the low concentrations of radioactivity in the critical plasma samples, the chances of success can be enhanced by employing good chromatographic practice. System refinements such as reducing dead volumes and shortening the distance between column and detector will increase the sensitivity, though it should be borne in mind that sharpening the chromatographic peak will shorten the residence time in the detector, potentially reducing the signal. Another consideration during the chromatographic analyses is the “quenching” effect that is often associated with radioactive measurements. Most investigators assume this to be constant throughout the chromatographic run although this is rarely checked. In addition to the radioactive components of interest, the samples will also contain a large number of co-eluting components present in the sample matrix. The presence or absence of quenching effects can be checked by running a blank sample and adding test material to the eluate. This can be achieved by a variety of methods such as direct infusion postcolumn, or postcolumn by collecting fractions and spiking.

During standard method establishment investigations, the suitability of the system is often established using parent material, but once a complex mixture is injected for analysis, the relative recovery of components can vary. The use of radiolabeled materials in the hADME study allows column recovery to be determined easily. This can alleviate concerns that material has been retained on the column, the stainless steel tubing, or the radiodetector cell. By measuring the column recovery, the system performance can be optimized as part of the method establishment process.

In an attempt to overcome the limitations associated with the radioflow systems, alternative methods were investigated and these can be divided into two main categories: a) stop-flow and b) dynamic flow methods. The stop-flow methods are as the name suggests based on stopping the flow once a radioactive peak is detected, thus providing longer detection times and therefore higher detection efficiencies. The peak is held within the detector as opposed to passing through therefore increasing the signal to noise ratio. The technology was originally developed for investigations performed in the agrochemical industry where samples routinely contain low levels of radioactivity. The advantages of stop-flow technology have been discussed in the literature (Nassar et al. 2004), but as yet the stop-flow technique has not been widely adopted in the drug development process. The major drawbacks associated with stop-flow technology are that subsequent analyses can result in inconsistent retention times and the detector is incompatible with LC/MS, meaning the samples need to be analyzed twice. Stop-flow has a lower throughput than alternative methods as counting occurs during the chromatographic run and therefore extends the chromatographic run times.

The use of a modified “dynamic flow” radiodetector was described by Cuyckens et al. (2008). Improvements in sensitivity were achieved by a modification to the standard online radiochemical detection system introducing the capability to provide variable scintillation fluid flow. Further improvements were achieved by reducing the internal diameter of the tubing, resulting in better peak shape, increased sensitivity, and higher resolution. When compared to conventional radio-HPLC using [3H]- and [14C]-labelled compounds, the method was reported to have comparable sensitivity to conventional techniques, was compatible with UPLC (thus shortening the chromatographic run times), and was suitable for hyphenation with mass spectrometers.

The detection limits obtained using the radio-chromatographic method can be significantly improved by performing the quantitation off-line. Eluent from the HPLC column can be fraction collected directly into scintillation vials, appropriate scintillation cocktail added, and the samples counted using standard liquid scintillation counting (LSC) methods. Each sample (fraction) can be counted for longer time periods, thus dramatically improving the sensitivity. On standard liquid scintillation counters, the samples are typically counted one vial at a time. If the chromatographic run is 30 min, with fractions collected at 15 s intervals, then counting each vial for 4 min would result in a total counting time of over 4 h.

The utility of off-line counting was significantly improved by the introduction of microplate scintillation counting (MSC) plates (Dear et al. 2006; Krauser et al. 2012). When performing chromatographic analysis with MSC counting, the eluent is collected directly into microplates, 96 or 384 well, using accurate fraction collectors. The technique has the advantage that several plates can be selected per run and the process can be automated thus improving the throughput. Two types of plate are commercially available, one with a solid scintillant base and another employing liquid scintillant. Use of solid scintillant plates results in a slightly lower sensitivity but allows recovery of notable fractions from the plate postcounting for further characterization using mass spectrometric techniques. Both types of plate give a notable improvement in sensitivity when directly compared to the results obtained using traditional fraction collection-liquid scintillation counting methods. Unlike traditional liquid scintillation counters, microplate scintillation counters are able to count multiple wells simultaneously (12–16 dependent on counter); therefore, the throughput is much higher. The technique should however be used with care as one of the key steps in sample analysis is evaporation of the eluent from the plate at which point there is the possibility to lose volatile components (parent drug or metabolites). It is therefore good practice to perform a system suitability check before committing the precious samples (especially plasma) for analysis. A good system suitability check would include a recovery check for radioactivity from plates spiked with parent compound and would ideally compare radioprofiles obtained from other biological matrices such as urine using both radioflow detection and microplate scintillation counting. An improved method for quantifying the radioactive content of microplate fractions has subsequently been reported (Dear et al. 2008). The method was essentially an imaging technique and was reported to shorten counting times required for analysis.

The counting methods detailed above were compared by Zhu et al. (2005) and the relative limits of detection discussed. The data are reproduced in Table 6 and show that of all the commonly used methods of radiometric detection the microplate scintillation counter provides the lowest limit of detection, with the exception of accelerator mass spectrometry. Whilst AMS has an extremely low detection, it is not a true radiometric method as quantitation is based on graphitization and measurement of [14C] atoms.
Table 6

Comparison radiodetector sensitivity

Radiodetection

Background(CPM)

Counting Efficiency (%)

Counting time (min)

Limit of Detection (DPM)

Limit of Quantification (DPM)

HPLC-RFD

15

70

5–10 s

250–500

750–1500

HPLC-LSC

25

90

10

10

31

HPLC-MSC

2

70

10

5

15

Stop-flow

15

70

1

25–50

75–150

HPLC-AMS

   

0.0001

 

Taken from Zhu et al. 2005

The relative sensitivity of the common techniques used for quantification of drug metabolites (radioflow, standard fraction collection, and microplate scintillation counting) was assessed in a comparative study using the same sample (McEwen et al. 2014). In general good agreement was observed using all three counting techniques, but overall the microplate scintillation counting provided greater resolution of the chromatographic peaks and a lower background. This consideration assumes great importance when selecting the analytical method for the analysis of clinical plasma samples where concentrations of drug-related material, and therefore concentrations of radioactivity, are generally low.

Alternatives to Carbon-14

NMR

The development of high field NMR machines in the 1980s led researchers to explore the use of NMR for analysis of plasma and urine samples. The technique was suitable for the study of both endogenous compounds (biomarkers) and drug metabolites. The technique requires limited sample preparation and is nondestructive, meaning that the sample can be retained for use in additional experiments.

The use of NMR for quantification and identification of drug metabolites in clinical and preclinical samples was developed by Nicholson (Nicholson et al. 1983, 1984a, 1985; Bales et al. 1984a, b, 1985). NMR is an inherently insensitive technique but has the advantage that rapid multicomponent analyses can be performed with limited sample preparation and with no prior assumptions as to the sample identity. The technique also provides structural information sometimes absent from mass spectrometric data such as the specific site of hydroxylation. Analyte concentrations need to be >50 μM for effective detection and the molecule requires suitable proton groups such as CH3-, -CH2-, or CH. The early promise faded as researchers turned to mass spectrometric techniques, but recently NMR has been used to generate quantitative data to answer regulatory questions (Dear et al. 2008; Caceres-Cortes and Reilly 2010). NMR has also been used to investigate likely routes of metabolism and provide an estimate of renal clearance (Nedderman et al. 2011) in the absence of traditional radioactive tracers.

Initial studies were conducted using proton NMR as most drugs contain hydrogen atoms, and this also provides structural information. For compounds containing fluorine, 19F NMR provides the opportunity to obtain quantitative data, although structural information obtained is limited. The low levels of fluorine in biological systems and the fact that fluorine 19 has a 100% isotopic abundance mean that all signals obtained can be related to drug or drug related material. This has led to many researchers using the technique in early drug metabolism (Dear et al 2000, Desmoulin et al. 2002, Ismail et al. 2002, Lenz et al. 2002, Malet-Martino et al. 2006). The technique continues to generate interest and recently two groups have published papers (James et al. 2017, Haitao et al. 2017) comparing the results obtained using F19 NMR with those obtained using the radiolabeled compounds.

Inductively Coupled Plasma Mass Spectrometry (ICP-MS)

ICP-MS is a technique that can detect a wide range of elements (metals and nonmetals) at extremely low concentrations, in the order of on part in 1015 (ppq). While the technique generally covers elements not found in the pharmaceutical compounds, for drugs containing halogen atoms, especially bromine and iodine, ICP-MS offers an alternative method for detection and quantification of drug-related material. The advantage of the technique is that detection involves atomization and ionization of the compound, meaning that quantification is independent of chemical structure and can be performed without the requirement for synthetic standards. The technique can be used to determine several elements simultaneously and can be combined with both normal and reverse phase HPLC to separate and quantify drug metabolites in biological fluids (Axelsson et al. 2001; Duckett et al. 2002). While ADME studies are currently not conducted for in the development of “biologicals,” it is interesting to note that the use of ICP-MS has recently been reported in the study of large molecule metabolism using a test compound labelled with iodine-127 (Lim et al 2014).

AMS

The concept of accelerator mass spectrometry (AMS) can be traced back to a cyclotron experiment conducted in the 1930s to measure 3H and 4He (Alvarez and Cornog 1939). The technique remained relatively underutilized until the publication of a paper by Richard A Muller in 1977. His paper in Science showed how particle accelerators (cyclotrons and linear) could be used for detection of tritium, carbon-14, and several other isotopes of scientific interest including beryllium-10, widely used in geology. He also published the first radiodating determination using tritium. Within month other groups (Nelson et al. 1977; Bennett et al. 1977) published further data using linear particle accelerators. These measurements were made using relatively large accelerators operating with terminal voltages of 7 and 8 MV. Since these early experiments, smaller instruments have been introduced and the trend toward smaller and smaller machines continues. Modern instruments operating at low accelerator voltages such as 200 kV for determination of 14C are being designed and built by companies such as ETH Zurich (Suter 2010).

Originally used in academic institutions as a radiocarbon dating technique , the ability to determine extremely low concentrations of radioactivity associated with carbon-14 and tritium was of great interest to scientists involved in the drug development process, especially those involved in the safety assessment of drug metabolites. The application of AMS in the drug development process has been the subject of several general reviews (Lappin and Garner 2005; Lappin and Stevens 2008). Despite being described in some quarters as a low level counting technique, AMS does not directly count radioactive decay particles but instead provides C12:C14 isotope ratios for the sample. AMS was first applied to biomedical studies in the 1980s (Litherland 1980). Since then further reviews detailing the biomedical applications of AMS have appeared (Lappin et al. 2006; Young and Ellis 2007; Young et al. 2008; Vogel et al. 2010; Seymour 2011). Specific examples of AMS investigations in the drug development process were also reported by Swart et al. (2016) and Bloomer et al. (2016).

One of the common misconceptions surrounding AMS is that it is a nonradioactive technique . While the amount of radioactivity involved in AMS studies is extremely low, such that the samples can be treated as nonradioactive and the usual human dosimetry is not required, the material administered to the volunteers is still radioactive and will require radiosynthesis in the usual way. Terms such as “lightly labelled” have been used, but the drug material should be prepared under accepted quality standards with a defined specific activity and structural characterization to ensure the study fulfils the regulatory objectives. While the regulatory requirements associated with administration of radioactive material to human volunteers may be eliminated, it is still good practice in the drug development process to obtain good quality data on the ADME properties of the investigational product in the preclinical species used in the toxicology assessments. The gold standard for these investigations is still the use of radiolabeled material. If these investigations are conducted around the same time as the hADME study, then contemporaneous analysis can be performed to compare the nature of circulating metabolites in preclinical and clinical samples.

While AMS provides exceptional limits of detection, it is not a technique that enables structural elucidation of the drug metabolites under analysis. Prior to analysis using AMS, all samples must be converted to elemental carbon, a process known as graphitization (Young et al. 2008). Samples of urine, fecal extracts, and/or plasma must be separated using chromatography, fractionated, and each fraction graphitized for analysis by AMS, thus generating a radioprofile similar to HPLC-LSC (Young and Ellis 2007; Young et al. 2008). Currently this is an inherently slow process requiring manual processing, and while progress in designing a system capable of transforming liquid inlet to a gaseous CO2 output has been reported (Liberman et al. 2004, 2007), this has yet to become commercially available.

Equally valuable information can be obtained using modern high resolution mass spectrometers using early clinical samples from the single ascending dose and multiple ascending dose (SAD/MAD) studies. In the hADME study, hyphenated LC-MS/MS-RAD analysis can help optimize the separation of components while obtaining good quality quantitative data.

It could be argued therefore that while AMS can undoubtedly provide a tool to solve specific problems associated with low bioavailability and/or extensive metabolism, adoption of AMS as the gold standard is still some way off and will ultimately depend upon further advances in AMS technology.

Summary

The hADME study is a pivotal study in the drug development process, providing a bridge between the preclinical toxicology findings and the clinical studies. The standard design involves administration of radiolabeled to human volunteers, collection of excreta and plasma, determination of a mass balance, and quantification and identification of the components in systemic circulation.

Examples

Example A: Dosimetry

One of the key activities in the planning of a hADME study is synthesis of the radiolabeled drug material, and this requires two pieces of information (1) the amount of radioactivity that can be administered to the volunteers and (2) the proposed chemical dose. The dosimetry study is therefore critical to providing this information forming the basis for the risk assessment in man.

Pigmented rats are used to assess binding of the drug material to melanin which is assessed from concentrations present in the uveal tract. Concentrations in pigmented skin can be measured, but there may be some confusion as to whether the observed radioactivity is bound to pigmented skin or the melanin present in the fur. An example image showing binding of test compound to the uveal tract is provided in Fig. 5.
Fig. 5

Binding to melanin – uveal tract

Tissue distribution studies in support of dosimetry calculations are generally designed to provide information on the distribution and kinetics of the radiolabeled material following dosing. A typical set of images obtained is provided in Fig. 6. In this example, it can be seen that the radioactivity distributes throughout the animal by 2 h following administration and is then seen to be eliminated through the kidney and gastrointestinal tract so that at 168 h following dosing radioactivity is seen only in the liver, caecum, and uveal tract. Inclusion of an additional timepoint at 504 h shows that radioactivity has been almost completely eliminated from the animal with the exception of the uveal tract. In the example, provided tissues (with the exception of the uveal tract) were clear of radioactivity by 504 h. As a general rule the earlier complete elimination is observed, the higher will be the radioactive dose that can be administered to man.
Fig. 6

Typical QWBA experiment for dosimetry assessment

To provide a realistic estimate of the risk associated with radioactivity present in the eye, the data can be analyzed to provide a biological half-life for modelling purposes. The tissue concentration data are provided Table 7 and the kinetic analysis is reproduced in Fig. 7. The data in this case indicate a half-life of around 213 h.
Table 7

Uveal tract data

Time (h)

Eye (uveal tract) concentration (μg equiv./g_

0

0

2

22.7

8

8.17

24

5.02

168

0.411

336

0.215

504

0.138

Fig. 7

Pharmacokinetic analysis of uveal tract data

One further factor affecting the result will be the relative routes of elimination of radioactivity. In this case, radioactivity was eliminated by both fecal and urinary routes. Dosimetry assessment for compounds showing high fecal elimination usually result in lower values than those with high urinary elimination. This is due to the longer residence time of radioactivity that passes through the gastrointestinal tract which results in greater internal exposure for this organ.

Example B: Mass Balance

A typical hADME study usually runs for a fixed period (nominally 7 days) with radioactive content in urine and feces measured to assess the mass balance recovery. It should be stressed that the primary objective is to assess the routes and rates of excretion of radioactivity and while a full mass balance is welcomed it is not always achievable. Daily measurement of the radioactivity excreted in urine and feces allows the elimination to be followed in real time and can provide useful information on the rates of elimination as well as predicting the time to reach 90%. A useful method is to plot “body burden graphs” as shown in the examples below.

Compound A

The hADME study for compound A was designed on the basis that the metabolism was well understood and a fixed term of 9 days was assigned for residence in the clinic. Urine and feces were collected on a daily basis and radioactive content determined. The elimination was then characterized using a body burden calculation as described earlier. The data obtained are provided in Table 8 and the kinetic analysis is shown in Fig. 8.
Table 8

Excretion data obtained for compound A

Time (h)

Urine

Feces

Body Burden (100 – (U+F))

0

0

0

100

24

52.0

2.08

45.9

48

13.8

3.38

28.8

72

4.66

2.90

21.2

96

2.16

1.13

17.9

120

1.11

1.47

15.3

144

0.66

0.60

14.1

168

0.50

0.26

13.3

192

0.29

0.18

12.8

216

0.24

0.10

12.5

Total

75.4

12.1

Fig. 8

Body burden compound A

Mean recovery obtained in this study was 87.5% (75.4% urinary, 12.1% fecal), thus below an arbitrary cut-off of 90% for a good recovery. The standard release criteria (<1% in two consecutive 24 h collections) were met in this study although there was no intention to extend collections beyond the original 9 days. The body burden analysis would have allowed a decision on whether to release the subjects and in this case the subjects would have been released anyway. By 9 days the elimination half-life was estimated at around 425 h (17 days) meaning that if 90% recovery was the sole criterion for release, the subjects would be confined to the clinic for a further 10–14 days, while if 95% had been set as the release criteria, then the clinical phase would have been extended by around 3 weeks. It can be argued that the elimination routes and rates of radioactivity have been well characterized in this study and that any extension would have added little to the overall conclusions.

Compound B

The hADME study for compound A was designed on the basis that the elimination of radioactivity (predominantly fecal) would be protracted and a fixed term of 17 days was assigned for residence in the clinic. Urine and feces were collected on a daily basis and radioactive content determined. The elimination was then characterized using a body burden calculation as described earlier. The data obtained are provided in Table 9 and the kinetic analysis is shown in Fig. 9.
Table 9

Excretion data obtained for compounds B and C

Time (h)

Compound B

Compound C

Urine

Feces

Body burden(100-(U+F)

Urine

Feces

Body burden(100-(U+F)

0

0

0

100

0

0

100

24

5.70

0.11

94.2

0.40

0.00

99.6

48

0.96

6.07

87.2

0.19

0.85

98.6

72

0.77

14.7

71.7

0.19

3.99

94.4

96

0.61

13.3

57.7

0.20

5.10

89.1

120

0.43

7.01

50.3

0.20

7.31

81.6

144

0.43

7.24

42.6

0.19

6.14

75.2

168

0.34

6.75

35.5

0.15

8.69

66.4

192

0.29

4.22

31.0

0.14

3.91

62.4

216

0.23

2.62

28.2

0.13

2.87

59.4

240

0.22

2.51

25.4

0.12

5.83

53.4

264

0.16

2.06

23.2

0.12

2.17

51.1

288

0.13

1.92

21.2

0.10

8.23

42.8

312

0.12

1.41

19.6

0.10

2.31

40.4

336

0.09

1.04

18.5

0.09

2.31

38.0

360

0.10

1.26

17.2

0.08

2.06

35.8

384

0.07

0.65

16.4

0.08

2.77

33.0

408

0.07

0.76

15.6

0.08

1.67

31.2

Total

10.7

73.7

2.56

66.2

Fig. 9

Body burden compound B

Mean recovery obtained in this study was 84.4% (10.7% urinary, 73.7% fecal) thus below an arbitrary cut-off of 90% for a good recovery. The standard release criteria (<1% in two consecutive 24 h collections) were met in this study although there was no intention to extend collections beyond the original 17 days. The body burden analysis would have allowed a decision on whether to release the subjects and in this case the subjects would have been released anyway. By 17 days the elimination half-life was estimated at around 272 h (11 days), meaning that if 90% recovery was the sole criterion for release the subjects would be confined to the clinic for a further 7–10 days, while if 95% had been set as the release criteria then the clinical phase would have been extended by around 2 weeks. As for compound A it can be argued that the elimination routes and rates of radioactivity have been well characterized in this study and that any extension would have added little to the overall conclusions.

Compound C

The hADME study for compound C was again designed on the basis that the elimination of radioactivity (predominantly fecal) would be protracted and a fixed term of 17 days was assigned for residence in the clinic. Urine and feces were collected on a daily basis and radioactive content determined. The elimination was then characterized using a body burden calculation as described earlier. The data obtained are provided in Table 9 and the kinetic analysis is shown in Fig. 10.
Fig. 10

Body burden graph compound C

Mean recovery obtained in this study was 68.8% (0.08% urinary, 66.2% fecal) thus below an arbitrary cut-off of 90% for a good recovery. The standard release criteria (<1% in two consecutive 24 h collections) were not met in this study although there was no intention to extend collections beyond the original 17 days. The body burden analysis would have allowed a decision on whether to release the subjects and in this case the subjects would have been confined to the clinic. By 17 days the elimination half-life was estimated at around 260 h (11 days), meaning that if 90% recovery was the sole criterion for release, the subjects would be confined to the clinic for a further 22 days, while if 95% had been set as the release criteria then the clinical phase would have been extended by around 4 weeks. As for compound A it can be argued that the elimination routes and rates of radioactivity have been well characterized in this study and that any extension would have added little to the overall conclusions.

Example C: Metabolite Quantification and Identification.

Once the radioactive measurements have been completed and a comparison of the AUC obtained from the bioanalytical measurements with the AUC obtained for total radioactivity performed, then the next key activity is the quantification and identification of notable metabolites.

The key to success at this stage is the preparation of a representative sample that can be analyzed using high performance liquid chromatography. Samples such as blood/plasma, bile, and urine are relatively easy to obtain and if they contain suitable quantities of radioactivity can be analyzed with minimal sample work up. Fecal samples require homogenization prior to analysis for radioactive content and can then be extracted prior to chromatography. The use of radiolabeled material is of benefit in this case as the efficiency of extraction can be determined and methods can be established beforehand using blank samples spiked with test compound. A limitation of the AMS technique is that the process is slower than if performed using standard radioactive doses. Fecal samples can be extracted, centrifuged and aliquots of the extract counted in real time, thus allowing decisions to be made following each set of counts.

Standard preparation methods usually employ concentration as the final step prior to analysis, and while losses can occur at every step, the potential for loss of material at this stage is quite high and should be assessed as part of the system suitability evaluation. A wide range of sample preparation methods are available; the most commonly employed are liquid-liquid extraction (Pedersen-Bjergaard and Rasmussen 2005) and solid-phase extraction (Moriwaki et al. 2002). The aim of the work up is to separate the analyte from the endogenous material, thus minimizing matrix effects and improving sensitivity (Kruve et al. 2009; Marchi et al. 2009). Ideally the sample preparation method should be as simple as possible, the greater the number of steps employed, the more likely it is that losses can occur or artifacts arise due to instability of the test compound.

As discussed earlier in the chapter, the standard approach for separation and quantification of drug metabolites is liquid chromatography followed by microplate scintillation counting. The utility of the technique will now be discussed. Prior to analysis of the plasma samples, the technique was checked for quench and evaporation.

Quench was checked by analyzing a blank plasma extract and collecting fractions into microplates. By use of a t-valve, a solution of parent compound could be introduced into the eluant prior to evaporation. The chromatographic analysis was performed using a 90-min gradient method with collection of 14 s fractions therefore requiring the use of four 96-well microsintillation plates. The results are shown in Fig. 11 and indicate the response is relatively constant over the first three plates. A spike in the detector response is sometimes obtained during plate changeover and analysis of results should take this into account. It is good practice therefore to perform analysis in duplicate to check for potential artefacts. Towards the end of the collection, there is a notable decrease in the response obtained, which then recovers to around the normal response by the end of the collection period. The reduction in response is caused by quenching due to endogenous material being washed off the column at high organic solvent ratio.
Fig. 11

Assessment of quench in the microplate counter

The effect of evaporation can be checked using cold material spiked into the injection solvent and loaded into the microplate wells. The solvent is then evaporated and the contents of the well checked using standard MS analysis. An example is provided in Fig. 12. Here solutions of parent and metabolite (M1) were placed into the microplate wells and the solvent removed using the standard method. Multiple wells were prepared for parent and metabolite and at different times the contents of one well were checked at using MS. In this example, it can be seen that the evaporative process has no effect upon the concentration of parent observed, but that after a certain time the concentration of M1 in the cells decreases probably due to insufficient organic solvent remaining, thus allowing the component to volatilize.
Fig. 12

Assessment of evaporation during sample processing

The relative sensitivities of the radioflow detector and the microplate counting method were discussed by Zhu et al. (2005) and a graphical example is provided in Fig. 13. In the upper chromatogram, a radioflow detector is presented and shows the presence of parent and a very noisy baseline. In the lower chromatogram, the same sample is analyzed using microplate scintillation counting and quite clearly shows the presence of two minor metabolites in circulation. All three components were quantified and the relative exposure established.
Fig. 13

Comparison radioflow and microplate detection methods

An additional advantage of the microplate technique is that given the long half-life of carbon-14 the samples can be re-counted to improve the signal to noise ratio. Figure 14 shows the data that can be obtained in this way. The upper chromatogram was reconstructed from fractions counted for 4 min. While there is a suggestion of some metabolites present in the sample, the situation can be resolved by extending the counting time to 30 min. In this chromatogram (lower trace), the presence of parent and four metabolites (M1–M4) can be established. Counting times of 30 min are generally uncommon, but in some cases plates have been analyzed for up to 50 min.
Fig. 14

Effect of extending counting time on microplate detector response

Given the widespread use of mass spectrometric techniques in the preclinical and early clinical phases of drug development, it is unsurprising that LC-MS/MS is the analytical method of choice for the identification of drug metabolites in the hADME study. There has been some debate within the bioanalytical community regarding methods for quantitation of circulating components, and in 2010 the European Bioanalysis Forum produced a white paper offering recommendations and discussing the issues (Timmerman et al. 2011).

The drawback associated with relying on LC-MS/MS alone for the quantification is that in the absence of standards the LC-MS response can show a 30-fold difference in relative exposure when compared to radiometric methods. The analyte response is structure and matrix dependent and subject to unpredictable ion suppression effects. By contrast, quantification using radiometric techniques is structure independent and less subject to matrix effects.

Having optimized the sample preparation methods and the chromatographic separation obtained, the final key decision is the choice of mass spectrometer used for identification. The mass analyzers of greatest use in the analysis of human plasma samples tend to be based on either Q-TOF (Mamyrin 2001) or Orbitrap (Erve et al. 2009) technology in part due to their high resolving power but also due to the associated data processing techniques. Commonly used data processing methods rely on background subtraction, neutral loss filtering, isotopic pattern recognition (and the use of 14C can introduce an isotope pattern into the molecule), and mass defect filtering.

As an example, a full scan mass spectrum obtained on analysis of plasma from a hADME study is provided in Fig. 15a. The main peaks seen here are not compound related, arising instead from endogenous compounds within the matrix. By applying a mass defect filter to the chromatogram, only compounds that are related to the test compound are detected and the spectrum has therefore been simplified (Fig. 15b). Further processing can then be directed onto relevant components and MS/MS experiments conducted. In this case, the addition of two oxygen atoms onto the parent structure has been confirmed (Fig. 15c) and the product spectrum can then be obtained for full identification of the drug metabolite (Fig. 15d).
Fig. 15

Use of data processing in metabolite ID investigations

References and Further Reading

  1. Alvarez L, Cornog R (1939) Helium and hydrogen of mass 3. Phys Rev 56:613CrossRefGoogle Scholar
  2. Attwood J, O’Brien B, Loaring HW, Williams A (2010) Evaluation of active counting mode option in LabLogic β-RAM model 5 and Laura software. Poster ISSX 2010. Istanbul, TurkeyGoogle Scholar
  3. Axelsson B-O, Jornten-Karlsson M, Michelsen P, Abou-Shakra F (2001) The potential of inductively coupled plasma mass spectrometry detection for high-performance liquid chromatography combined with accurate mass measurement of organic pharmaceutical compounds. Rapid Commun Mass Sprectrom 15:375–385CrossRefGoogle Scholar
  4. Baillie TA, Cayen MN, Fouda H, Gerson RJ, Green JD, Grossman SJ, Klunk LJ, LeBlanc B, Perkins DG, Shipley LA (2002) Drug metabolites in safety testing. Toxicol Appl Pharmacol 182:188–196CrossRefPubMedGoogle Scholar
  5. Bales JR, Higham DP, Howe I, Nicholson JK, Sadler PJ (1984a) Use of high-resolution proton nuclear magnetic resonance spectroscopy for rapid multi-component analysis of urine. Clin Chem 30:426Google Scholar
  6. Bales JR, Sadler PJ, Nicholson JK, Timbrell JA (1984b) Urinary excretion of acetaminophen and its metabolites as studied by proton NMR spectroscopy. Clin Chem 30:1631Google Scholar
  7. Bales JR, Nicholson JK, Sadler PJ (1985) Two-dimensional proton nuclear magnetic resonance “maps” of acetaminophen metabolites in human urine. Clin Chem 31:757Google Scholar
  8. Bennett CL, Beukens RP, Clover MR, Gove HE, Liebert RB, Litherland AE, Purser KH, Sondheim WE (1977) Radiocarbon dating using electrostatic accelerators: negative ions provide the key. Science 198:508–510CrossRefPubMedGoogle Scholar
  9. Bloomer JC, Beaumont C, Dear GJ, North S, Young G (2016) The value of metabolite identification and quantification in clinical studies. Some case studies enabling early assessment of safety in humans: GlaxoSmithKline. In: Metabolite safety in drug development. Wiley, Hoboken, NJ, pp 275–292CrossRefGoogle Scholar
  10. Caceres-Cortes J, Reilly MD (2010) NMR spectroscopy as a tool to close the gap on metabolite characterisation under MIST. Bioanalysis 2:1263–1276CrossRefPubMedGoogle Scholar
  11. Cuyckens F, Koppen V, Kembuegler R, Leclercq L (2008) Improved liquid chromatography-Online radioactivity detection for metabolite profiling. J Chromatogr A 1209:128–135CrossRefPubMedGoogle Scholar
  12. Dear GJ, James AD, Sarda S (2006) Ultra-performance liquid chromatography coupled to linear ion trap mass spectrometry for the identification of drug metabolites in biological samples. Rapid Commun Mass Spectrom 20:1351–1360Google Scholar
  13. Dear GJ, Roberts AD, Beamont C, North SE (2008) Evaluation of preparative high performance liquid chromatography and cryoprobe nuclear magnetic resonance spectroscopy for the early quantitative estimation of drug metabolites in early clinical studies. J Chromatogr B 876:182–190CrossRefGoogle Scholar
  14. Desmoulin F, Gilard V, Malet-Martino M, Martino R (2002) Metabolism of capecitabine, an oral fluorouracil prodrug: 19F NMR studies in animal models and human urine. Drug Metab Dispos 30:1221–1229CrossRefPubMedGoogle Scholar
  15. Duckett CJ, Bailey NJC, Walker H, Abou-Shakra F, Wilson ID, Lindon JC, Nicholson JK (2002) Quantitation in gradient high performance liquid chromatography/inductively coupled mass spectrometry using diclofenac and chlorpromazine. Rapid Commun Mass Spectrom 16:245–247CrossRefPubMedGoogle Scholar
  16. EMEA (2012) ICH guideline M3 (R2) – questions and answers EMA/CHMP/ICH/507008/2011Google Scholar
  17. Erve JCL, Gu M, Wang Y, DeMaio W, Talaat RE (2009) Spectral accuracy of molecular ions in an LTQ/Orbitrap mass spectrometer and implications for elemental composition determination. J Am Soc Mass Spectrom 20:2058–2069CrossRefPubMedGoogle Scholar
  18. European Clinical Trials Directive Communication from the Commission – Detailed guidance on the request to the competent authorities for authorisation of a clinical trial on a medicinal product for human use, the notification of substantial amendments and the declaration of the end of the trial (CT-1) (2010/C 82/01) Available at https://ec.europa.eu/health/sites/health/files/files/eudralex/vol-10/2010_c82_01/2010_c82_01_en.pdf Accessed 8 Aug 2017
  19. Haitao Hu, Kishore Kumar Katyayan, Boris A. Czeskis, Everett J (2017) Perkins and Palaniappan Kulanthaivel Drug Metabolism and Disposition 45(4):399–408Google Scholar
  20. Hamilton RA, Garnett WR, Kline BJ (1981) Determination of mean valproic acid serum level by assay of a single pooled sample. Clin Pharmacol Ther 29:408–413CrossRefPubMedGoogle Scholar
  21. Hill H (2009) Development of bioanalysis: a short history. Bioanalysis 1(1):3–7CrossRefPubMedGoogle Scholar
  22. Hop CECA, Wang Z, Chen Q, Kwei G (1998) Plasma-pooling methods to increase throughput for in vivo pharmacokinetic screening. J Pharm Sci 87(7):901–903CrossRefPubMedGoogle Scholar
  23. Hu H, Katyayan KK, Czeskis BA, Kulanthaivel EJPP (2017) Comparison between radioanalysis and 19f nuclear magnetic resonance spectroscopy in the determination of mass balance , metabolism, and distribution of pefloxacin. Drug Metab Dispos 45(4):399–408CrossRefPubMedGoogle Scholar
  24. Humphreys WG, Unger SE (2006) Safety assessment of drug metabolites: characterisation of chemically stable metabolites. Chem Res Toxicol 19:1564–1569CrossRefPubMedGoogle Scholar
  25. ICH (2009) M3 (R2) non-clinical safety studies for the conduct of human clinical trials and marketing authorisation for pharmaceuticals. Center for Drug Evaluation and Research (CBER), Silver Spring, MD, USAGoogle Scholar
  26. ICRP (1992) International commision on radiological protection. Radiological protection in biomedical research. ICRP publication 62 Ann ICRP 22 No 3Google Scholar
  27. Ismail IM, Dear GJ, Roberts AD, Plumb RS, Ayrton J, Sweatma BC, Bowers GD (2002) N-O glucuronidation: a major human metabolic pathway in the elimination of two novel anti-convulsant drug candidates. Xenobiotica 32:29–43CrossRefPubMedGoogle Scholar
  28. James AD, Marvalin C, Luneau A, Meissner A, Camenisch G (2017) Comparison of 19F NMR and 14C measurements for the assessment of ADME of BYL719 (Alpelisib) in humans. Drug Metab Dispos 45:900–907CrossRefPubMedGoogle Scholar
  29. Kuerzel U, Krone V, Zimmer M, Shackleton G (2011) The human ADME study. Drug discovery and evaluation: methods in clinical pharmacology, pp 73–103Google Scholar
  30. Krauser J, Walles M, Wolf T, Graf D, Swart P (2012) A unique automation platform for measuring low level radioactivity in metabolite identification studies. PLoS One 7:e39070CrossRefPubMedPubMedCentralGoogle Scholar
  31. Kruve A, Leito I, Herodes K (2009) Combating matrix effects in LC/ESI/MS: the extrapolative dilution approach. Anal Chim Acta 651(1):75–80CrossRefPubMedGoogle Scholar
  32. Lappin G, Garner RC (2005) The use of accelerator mass spectrometry to obtain early human ADME/PK data. Expert Opin Drug Metab Toxicol 1:23–31CrossRefPubMedGoogle Scholar
  33. Lappin G, Stevens L (2008) Biomedical accelerator mass spectrometry: recent applications in metabolism and pharmacokinetics. Expert Opin Drug Metab Toxicol 4:1021–1033CrossRefPubMedGoogle Scholar
  34. Lappin G, Rowland M, Garner RC (2006) The use of isotopes in the determination of absolute bioavailability of drugs in humans. Expert Opin Drug Metab Toxicol 2:419–427CrossRefPubMedGoogle Scholar
  35. Leblanc B, Jezequel S, Davies T, Hanton G, Taradach T (1998) Binding of drugs to eye melanin is not predictive of ocular toxicity. Regul Toxicol Pharmacol 28:124–132CrossRefPubMedGoogle Scholar
  36. Lenz EM, Wilson ID, Wright B, Partridge EA, Rodgers CT, Haycock PR, Lindon JC, Nicholson JK (2002) A comparison of quantitative NMR and radiolabelling studies of the metabolism and excretion of Statil (3-(4-bromo-2-fluorobenzyl)-4-oxo-3H-phthalazin-1-ylacetic acid) in the rat. J Pharm Biomed Anal 28:31–43CrossRefPubMedGoogle Scholar
  37. Liberman RG, Tannenbaum SR, Hughey BJ, Shefer RE, Klinkowstein RE, Prakash C, Harriman SP, Skipper PL (2004) An interface for direct analysis of 14C in nonvolatile samples by accelerator mass spectrometry. Anal Chem 76:328–334CrossRefPubMedGoogle Scholar
  38. Liberman RG, Skipper PL, Prakash C, Shaffer CL, Flarakos J, Tannenbaum SR (2007) BEAMS Lab: novel approaches to finding a balance between throughput and sensitivity. Nucl Instr Meth Phys Res B 259:773–778CrossRefGoogle Scholar
  39. Lim H-K, Yuan C, Qiu X, Silva J, Evans DCA (2015) non-radioactive approach to investigate the metabolism of therapeutic peptides by tagging with 127I and using ICP-MS analysis. Drug Metab Dispos 43(1):17–26.  https://doi.org/10.1124//dmd.114.059774 CrossRefPubMedGoogle Scholar
  40. Litherland AE (1980) Ultrasensitive mass spectrometry with accelerators. Am Rev Nucl Part Sci 30:437–473CrossRefGoogle Scholar
  41. Lockley WJS, McEwen A, Cooke R (2012) Tritium: a coming of age for drug discovery and development ADME studies. J Label Compd Radiopharm 55:235–257CrossRefGoogle Scholar
  42. Malet-Martino M, Gilard V, Desmoulin F, Martino R (2006) Fluorine nuclear magnetic resonance spectroscopy of human biofluids in the field of metabolic studies of anticancer and antifungal fluoropyrimidine drugs. Clin Chim Acta 366:61–73CrossRefPubMedGoogle Scholar
  43. Mamyrin BA (2001) Time-of-flight mass spectrometry (concepts, achievements, and prospects). Int J Mass Spectrom 206(3):251–266CrossRefGoogle Scholar
  44. Marchi I, Rudaz S, Veuthey JL (2009) Sample preparation development and matrix effects evaluation for multianalyte determination in urine. J Pharm Biomed Anal 49(2):459–467CrossRefPubMedGoogle Scholar
  45. McEwen A (2011) Low level radioactive counting. In: Scientific presentation CropWorld London 31 October 2011Google Scholar
  46. McEwen A, Wood SG, Wilson KH, Ford GR (2012) Human ADME studies: a post MIST analysis. In: Poster presentation 12th European ISSX Meeting 17–21 June 2012, the NetherlandsGoogle Scholar
  47. McEwen A, Henson CM, Wood SG (2014) Quantitative whole-body autoradiography, LC-MS/MS and MALDI for drug-distribution studies in biological samples: the ultimate matrix trilogy. Bioanalysis 6(3):377–391; Miller R (1977) Radioisotope dating with a cyclotron. Science 196(4289):489–494Google Scholar
  48. Moriwaki H, Watanabe A, Arakawa R, Tsujimoto Y, Shimizu M, Noda T, Warashina M, Tanaka M (2002) Simultaneous determination of metabolites of trimethylbenzenes, dimethylbenzylmercapturicacid and dimethylhippuric acid, in human urine by solid-phase extraction followed by liquid chromatography tandem mass spectrometry. J Mass Spectrom 37:1152–1157CrossRefPubMedGoogle Scholar
  49. Naito S, Furuta S, Yoshida T, KitadaO O, Fueki T, Unno Y, Ohno H, Onodera N, Kaeamura M, Kurokawa F, Sagami K, Shinoda T, Nakazawa T, Yamazak T (2007) Current opinion: Safety evaluation of drug metabolites in development of pharmaceuticals. J Toxicol Sci 32:329–341CrossRefPubMedGoogle Scholar
  50. Nassar A-EF, Parmentier Y, Martinet M, Lee DY (2004) Liquid chromatography–accurate radioisotope counting and microplate scintillation counter technologies in drug metabolism studies. J Chromatogr Sci 42:348–353CrossRefPubMedGoogle Scholar
  51. Nedderman AN, Dear GJ, North S, Obach RS, Higton D (2011) From definition to implementation: a cross industry perspective of past, current and future MIST strategies. Xenobiotica 41:605–622CrossRefPubMedGoogle Scholar
  52. Nelson DE, Korteling RG, Stott WR (1977) Carbon-14: direct detection at natural concentrations. Science 198:507–508CrossRefPubMedGoogle Scholar
  53. Nicholson JK, Buckingham MJ, Sadler PJ (1983) High resolution 1H n.m.r. studies of vertebrate blood and plasma. Biochem J 211(3):605Google Scholar
  54. Nicholson JK, Bales JR, Sadler PJ, Juul SM, Mcleod A, Sonksen PH (1984a) Monitoring metabolic disease by proton NMR of urine. The Lancet ii:751Google Scholar
  55. Nicholson JK, Flynn MO, Sadler PJ, Macleod A, Juul SM, Sonkson PH (1984b) Proton-nuclear-magnetic-resonance studies of serum, plasma and urine from fasting normal and diabetic subjects. Biochem J 217:365Google Scholar
  56. Nicholson JK, Timbrell J, Sadler PJ (1985) Two-dimensional proton nuclear magnetic resonance “maps” of acetaminophen metabolites in human urine. Mol Pharmacol 27:644Google Scholar
  57. Pedersen-Bjergaard S, Rasmussen KE (2005) Bioanalysis of drugs by liquid microextraction coupled to separation techniques. J Chromatogr B 817:3–12CrossRefGoogle Scholar
  58. Prueksaritanont T, Lon JH, Baillie T (2006) Complicating factors in safety testing of drug metabolites: kinetic differences between generated and preformed metabolites. Toxicol Appl Pharmacol 217:143–152CrossRefPubMedGoogle Scholar
  59. Seymour MA (2011) Accelerator MS: its role as a frontline bioanalytical technique. Bioanalysis 3:2817–2823CrossRefPubMedGoogle Scholar
  60. Smith DA, Obach RS (2005) Seeing through the MIST: abundance versus percentage. Commentary on metabolites in safety testing. Drug Metab Dispos 33:1409–1417CrossRefPubMedGoogle Scholar
  61. Smith DA, Obach RS (2006) Metabolites and safety: what are the concerns and how should we address them? Chem Res Toxicol 19:1570–1579CrossRefPubMedGoogle Scholar
  62. Suter M (2010) Challenging developments in three decades of accelerator mass spectrometry at ETH: from large particle accelerators to table size instruments. Eur J Mass Spectrometry 16(3):471–478.  https://doi.org/10.1255/ejms.1078 CrossRefGoogle Scholar
  63. Swart P, Lozac’h F, Zollinger M (2016) A mass balance study and metabolite profiling of sonidegib in healthy male subjects using the microtrace approach. In: Metabolite safety in drug development. Wiley, Hoboken, NJ, pp 243–260CrossRefGoogle Scholar
  64. Timmerman P, Kall MA, Gordon B, Laakso S, Freislehen A, Hucker R (2011) Bioanalysis 2:1185–1194CrossRefGoogle Scholar
  65. Ullberg S (1954) Studies on the distribution and fate of 35S-labelled benzylpenicillin in the body. Acta Radiol Suppl 118:1–110PubMedGoogle Scholar
  66. Ure A (1841) Gouty concretions. Pharm J Transact 1:30–35Google Scholar
  67. US FDA guidance for industry (2008) Safety testing of drug metabolitesGoogle Scholar
  68. US FDA guidance for industry (2016) Safety testing of drug metabolitesGoogle Scholar
  69. Vogel JS, Giacomo JA, Schulze-König T, Keck BD, Lohstroh P, Dueker S (2010) Accelerator mass spectrometry best practices for accuracy and precision in bioanalytical 14C measurements. Bioanalysis 2:455–468CrossRefPubMedGoogle Scholar
  70. Woehler F, Tiedemann F (1824) Z Physiol 1:14Google Scholar
  71. World Health Organisation (1977) Use of ionising radiation and radionuclides on human beings for medical purposes. Technical report No 611, GenevaGoogle Scholar
  72. Young GC, Ellis WJ (2007) AMS in drug development at GSK. Nucl Inst Methods Phys Res B 259:752–757CrossRefGoogle Scholar
  73. Young GC, Corless S, Felgate CC, Colthup PV (2008) Comparison of a 250 kV single-stage accelerator mass spectrometer with a 5 MV tandem accelerator mass spectrometer-fitness for purpose in bioanalysis. Rapid Commun Mass Spectrom 22:4035–4042CrossRefPubMedGoogle Scholar
  74. Zhu M, Zhao W, Vazquez N, Mitroka JG (2005) Analysis of low level radioactive metabolites in biological fluids using high-performance liquid chromatography with microplate scintillation counting: method validation and application. J Pharm Biomed Anal 39:233–245CrossRefPubMedGoogle Scholar

Copyright information

© Springer International Publishing AG 2017

Authors and Affiliations

  1. 1.Reverisco SolutionsPeterboroughUK

Personalised recommendations