Developability:  A Science for Discovery Lead Candidate Selection

Duane B. Lakings, Ph.D., Principal, DSE Consulting

Preliminary in vivo experiments can assess the developability of discovery leads for human use.  Positive results allow earlier and more cost-effective entry into clinical trials with well-designed, data-productive clinical protocols.

Your completed discovery experiments suggest a compound, or class of compounds, mediates a disease process and has human therapeutic potential.  Is the lead now ready to be transferred from discovery to development?  Or should additional studies be conducted?  If so, what experiments or studies should be done? 

This DSE Consulting White Paper describes some of the biological research studies that could, and in most cases should, be conducted to determine if a discovery lead has the physical and chemical, as well as the biological, properties to be a drug candidate.  These developability studies may also uncover concerns that must be resolved before beginning the definitive nonclinical studies needed to support an IND filing and before designing the clinical protocols, which will evaluate the candidate's safety and efficacy in humans.  Identifying and correcting these concerns as early as possible allows for a more timely and data productive development program.  A program with substantially fewer costly repeat studies, in terms of both time and money, and one with a much better chance of successfully completing Phase III clinical testing.

Development Questions

Before initiating a development program, a number of important questions should be asked and the best-guess answers sought to plan the research studies and the timelines for their completion.  These questions include, but are not limited to, the following:

•What is the disease indication for the candidate?

•What is the purposed clinical route and frequency of dosing?

•What is the estimated candidate concentration in a physiological fluid and how long should that concentration be maintained to obtain the desired pharmacological response?

•What, if any, are the biological markers to monitor toxicity or therapeutic effectiveness?

•What nonclinical studies need to be completed before initiating the clinical program?

•What is the projected timeline for the nonclinical studies and for filing the IND?

Six scientific disciplines are involved in this early compound characterization and are  1)  analytical and bioanalytical,  2)  pharmacology,  3)  nonclinical formulation;  4)  pharmacokinetics (PK),  5)  drug metabolism (DM), and  6)  pathology and toxicology (Path/Tox).  Each of these disciplines is described in more detail below.

Analytical and Bioanalytical Chemistry Methods

If not already available, an assay(s) should be defined for the quantification of the lead candidate in physiological fluids and in formulations.  This method(s) can support experiments in pharmacology, formulation development, PK, DM, and Path/Tox.  For preliminary studies, the assay should be characterized to demonstrate the range of reliable results (RRR), the lower (LLQ) and upper (ULQ) limits of quantification, specificity, accuracy, and precision.  In addition, evaluations on the matrix to be used (blood, plasma, serum) should be conducted and the stability of the lead candidate in the selected matrix determined.  For nonclinical formulations, the assay should be stability indicating so that it can predict the candidate's degradation profile. 

The first step in characterizing an assay is to select the analytical technique.  For a small organic molecule, HPLC or GC may be employed.  A macromolecule may require an ELISA or RIA method.  Samples in assay diluent and in matrix and over a large concentration range are analyzed to assess if the technique produces an appropriate signal and to determine assay specificity, which is the presence of potential interferences from matrix components and from the different animal species to be used in pharmacology and Path/Tox experiments.  RRR and reliability of quantification are assessed from standard curves and multiple samples fortified at two or more concentrations.  The standard curve responses that can be described by a mathematical equation (linear, quadratic, sigmoidal) define RRR.  LLQ is the lowest signal that can be accurately measured above background and ULQ is the highest signal defined by the response curve.  The fortified sample results determine the precision and accuracy of the assay.  The ability to measure a compound in a physiological fluid is not useful if the lead candidate is unstable.  Conducting a preliminary stability study ensures that the lead does not degrade in blood, during processing to obtain plasma or serum, during the time and under the conditions that specimens may be stored until analyzed, and during sample preparation. 

A stability-indicating assay predicts whether the lead candidate in a formulation degrades from the time of preparation to the time of dosing.  The lead's physical and chemical properties usually indicate a technique (heat, light, temperature, pH) that can degrade the compound.  Samples, stored under both nondegrading and degrading conditions for various lengths of time, are assayed and the results evaluated for changes in the lead candidate's concentration over time. 

Successful completion of the above experiments will characterize assays for use in evaluating a candidate in animal models and in nonclinical formulations.  Once a candidate is selected for further development, the methods need to be validated for each matrix and for each species before being utilized to support definitive GLP regulated Path/Tox, DM, and PK studies.

Pharmacology Studies

Preliminary pharmacological evaluations will have shown that a discovery lead interacts with a biological process that may be of therapeutic benefit.  Depending on the design and extent of these early studies, additional pharmacology studies in appropriately defined animal models may be needed to further characterize the dose response curve using the proposed clinical route and frequency of dosing.  If possible, these studies should be conducted in at least two species and the dose that gives the desired biological response should be determined.  This dose level, divided into the no-toxic effect dose in the same species, calculates a therapeutic ratio (TR).  If the TR is one or less, a lead will elicit adverse events (AEs) in addition to the beneficial response.  Unless the candidate is for the treatment of a life-threatening disease such as AIDS or certain CNS indications, a low TR is a warning sign that the lead may not have the necessary properties for further development.  A TR of five or greater indicates the lead will most likely produce a pharmacological response before causing dose-limiting toxicity.  If possible, these pharmacology studies should be conducted with dosing to steady state.  The number of doses required to reach steady state depends on the PK profile in the same animal model.  These multiple-dose studies provide information on the frequency of dosing necessary to maximize the biological response.

Pharmacology evaluations assist in the selection of dose levels and route and frequency of dosing for preliminary and definitive Path/Tox studies and for Phase I human studies.  If the effective pharmacological dose is unknown, underdosing and achieving no therapeutic response or overdosing and being unable to define a no-toxic effect dose are equally possible.  In such cases, the development of a potential beneficial therapeutic agent could be discontinued.

Nonclinical Formulation Studies

Nonclinical formulation definition and the drug delivery characteristics of discovery leads are not usually studied in detail during the transition from discovery to development.  The definition of an acceptable formulation depends on the proposed route of administration.  For IV administration, compatibility with blood is necessary so that the lead does not precipitate and has minimal local toxicity.  Compounds that are highly lipophilic or have limited aqueous solubility are the most likely to encounter these types of problems.  A low extent of or a high variability in absorption can cause problems for a lead administered by other routes (PO, SC, dermal).  For a poorly absorbed candidate, the amount reaching the site of action may be insufficient to elicit or maintain the desired biological response.  If the absorption is variable and the TR is low, a toxic response may be observed in some animals and later in humans, if the candidate reaches clinical testing.  Generally, when the extent of absorption is ≥50% and the variability of absorption is <50% of the amount absorbed, the lead has acceptable bioavailability for further development.  For a candidate with an extent of absorption <25% of the administered dose or a variability of absorption of >100% of the amount absorbed, other formulations with absorption enhancers or solubilizers might be evaluated to improve the drug delivery profile.  If improvement is not possible, the chances of a lead with low, variable absorption becoming a therapeutic product are reduced, and the candidacy of such a compound should be carefully considered.

A stability-indicating assay is used to determine the amount of compound in formulations used for dosing animals in PK, DM, and Path/Tox studies.  Results from these analyses ensure that the formulations contain the desired amount of compound, that the concentration did not change during the dosing period, and that the animals are receiving the desired dose levels.

Without an acceptable formulation, the extent and variability of delivery may make interpretation of results from other developability studies meaningless and prevent the continued development of a potentially useful therapeutic agent.

Animal Pharmacokinetics

The first animal PK study confirms that the bioanalytical method can characterize the absorption and disposition profiles of the lead candidate.  A study design for a lead that has pharmacological activity when administered PO to rats may consist of dosing four rats: two with IV bolus injections at a dose level between 25 and 50% of the pharmacological active dose and two PO at the pharmacological active dose.  The plasma-concentration-versus-time profiles after IV dosing provide information on the distribution and disposition kinetics.  These IV results also define the concentration range that can be expected in animals and assist in determining the sampling times to be used in more definitive animal PK experiments.  The plasma concentration profiles after PO dosing provide information on the absorption kinetics and on the absolute bioavailability of the lead.

The design of additional animal PK studies depends on the results of the preliminary animal PK study, the theoretic kinetic profile needed to produce the desired pharmacology, the results from preliminary Path/Tox experiments, and the goals of the drug development program.  Carefully designed and conducted experiments optimize the chance of obtaining the desired information.  If the results show a high plasma concentration some hours after either IV or non-IV dosing, suggesting a well absorbed, slowly clearly compound, dosing, additional PK studies should be designed to confirm these data and evaluate dose proportionality and the linearity of kinetics over the dose range used in preliminary Path/Tox studies.  If the results indicate a low plasma concentration shortly after IV dosing, additional PK studies should demonstrate that this profile provides the desired pharmacology, that more frequent dosing does or does not change the pharmacological profile, and whether the low concentration is caused by rapid clearance or by distribution into storage compartments.  If a lead is extensively metabolized, preliminary DM studies become important since the metabolites may have pharmacological or toxicological activity and the metabolites most likely will have a PK profile different from the parent compound.  If a lead is cleared by excretion, the organs of elimination will be subjected to the highest concentrations of a compound and are the most likely organs to show signs of toxicity.  For a lead with good bioavailability but rapid clearance, additional studies would be similar to those described above for a lead administered IV and with a rapid clearance.  If a lead has a low F but was slowly cleared so that the desired concentration was reached and maintained for the desired time, additional PK studies should determine the cause for the low absorption, show that the low absorption was or was not dose dependent, and evaluate the extent of inter-animal variability.  If the low absorption is caused by poor biopharmaceutical properties, additional formulations can be tested to overcome the problem.  A new formulation would also need to be evaluated in the animal pharmacology model(s) to determine the extent of change in the dose-response curve.  If the preliminary Path/Tox experiments have already been conducted, these safety studies should be repeated with the new, more available formulation to determine changes in the toxicity profile.

For many drug development programs, Path/Tox studies in two or more species are necessary.  In this case, preliminary animal PK studies should be conducted in each species used in animal safety studies.  If differences in delivery or disposition exist between species and result in an enhanced or decreased toxicology profile, PK results may explain, in part, the different toxicology profiles.  If possible, physiological fluid specimens should be obtained from animals in the preliminary Path/Tox studies to determine the extent of exposure or toxicokinetics (TK).  The results provide information on possible changes in exposure and on the accumulation potential of the compound and can be used to design multiple-dose animal PK and tissue distribution studies.  If the change in disposition or accumulation is substantial, modification of the dosing regimen may be necessary to obtain the desired concentration profile after dosing to steady state.

Drug Metabolism

The number and design of preliminary DM studies during the transition period depend on the results from preliminary PK and Path/Tox studies.  DM evaluations determine distribution, disposition, metabolism, and elimination or ADME and are usually conducted with a radiolabeled compound.  If a lead has a slow disposition phase, suggesting distribution into some extravascular tissues, or if preliminary Path/Tox experiments identify potential organs of toxicity, a preliminary tissue distribution (TD) and mass balance (MB) study can be designed.  This study can evaluate the distribution of drug-related material into selected tissues, such as liver, kidney, fat (for lipophilic compounds), muscle, skin, heart, and determine the primary route(s) and rate(s) of elimination.  The total radioactivity minus the parent compound concentration estimates the amount of metabolites present.  If the difference is minimal and does not change over time, the extent of metabolism is low.  For bile or urine specimens, high levels of radioactivity and small differences between radioactivity level and parent compound concentration indicate a primary route of elimination for the parent with little metabolism.  If the results from other animal studies suggest potential accumulation in various organs, a preliminary multipledose TD study can be conducted with dosing to steady state.  For a lead cleared primarily by metabolism, a metabolite profile in urine, bile, or air can determine the number of potential metabolites.  When the level of a metabolite in a matrix is high, attempts to isolate and identify the metabolite(s) can be undertaken and its (their) pharmacological and toxicological activity can be evaluated, providing information on the mechanism of action for the compound. 

Once a lead has been selected for development, the results from these preliminary DM studies can be used to assist in the design of the definitive Path/Tox and DM studies, which are conducted according to GLP guidelines.

Pathology and Toxicology

Path/Tox studies are conducted to define the drug safety profile of a candidate that includes the no-toxic-effect dose, maximal tolerated dose, potential organs of toxicity, and potential biochemical markers to detect and track toxic events.  Most developmental compounds that do not become therapeutic products have unacceptable toxicity in animals or in humans.  Before initiating the definitive Path/Tox studies needed to support an IND application, a number of experiments can be conducted to characterize the potential toxicity of a lead in animals.

To evaluate the qualitative and quantitative single-dose toxicity of a lead, a single dose at a number of dose levels is administered by the proposed clinical route and the animals, usually rats or mice, are observed for 14 days.  The number of animals in each dose group and the number of dose levels to be evaluated depends on the pharmacology profile of the lead candidate and the expected toxicity profile, usually obtained from toxicity studies on related compounds.  This study can also assess local irritation at the site of administration.

The doses for definitive GLP Path/Tox studies are defined in dose-range-finding studies.  These studies usually include four dose levels, the highest level being the highest dose that did not cause acute toxicity, and a control and usually have 6-10 animals (3-5 per sex) in a dose group.  Endpoints for dose-range-finding studies may include weight loss, activity changes, clinical chemistry changes, and histology and pathology evaluations at necropsy.  The primary goal is to determine a maximum-tolerated dose using the route and frequency of administration proposed for clinical testing.

A dose level that causes toxic changes (morbidity, salivation) and one that produces the no toxic effect are determined during 14day studies.  For a candidate to be used for a nonlifethreatening clinical indication, at least two animal species are tested, one rodent and one nonrodent.  The information gained is used to model the definitive, GLP Path/Tox studies so that these are conducted with a cost-efficient design and are data productive.  These pilot studies also can evaluate the potential for antibody production and clinical chemistry changes in physiological parameters.  These data identify potential biological markers that can be employed in other studies to evaluate, and possibly predict, adverse events.

Organs of toxicity may be identified by a full histological workup of animals in each dose group and from the results obtained from the analyses of clinical chemistry samples.  The candidate levels in the identified organs of toxicity can be determined by the DM group in an attempt to correlate the observed toxicity with high or accumulated concentrations (toxicodynamic correlation).  If possible, investigations into the biochemical mechanism of toxicity should be initiated.  If results from the preliminary Path/Tox studies show a lead has an unacceptable level of toxicity, the development candidacy of such a compound should be carefully considered.

From Discovery to Development

This white paper describes some preliminary experiments in six scientific disciplines that can characterize the developability of a lead discovery candidate before the lead enters the definitive nonclinical and clinical drug development processes.  With appropriate planning and commitment of resources, these studies can usually be completed in three to six months, if major problems are not encountered in one or more of the scientific areas.  If these experiments are completed as part of the transition from discovery to development, compounds that do not have the characteristics necessary to become therapeutic agents can be identified early and prevented from entering the development process.  Analogues of a compound with unacceptable characteristics can be evaluated to find a development candidate that has more optimal properties.  In addition, the results from the preliminary studies will allow the definitive nonclinical development studies to be designed and conducted in a timely, costefficient manner, and thus allow the candidate to have an earlier entry into the clinic.  A well-characterized drug candidate will have a better change of successfully completing the clinical program and becoming a therapeutic product.