Clinical development of biological response modifiers

OBJECTIVE: To present perspectives on selected issues that frequently arise during the clinical development of biological response modifier (BRM) therapies.


us
B IOLOGICAL RESPONSE MOOIF'IER (BRM) Tl IERAPY RE PERS TO therapeu lie interventions that t.arget immunological.inflammatory or oilier host responses to infection, malignancy or other abnormal physiological events.The pace of development of BRM therapy has increased exponentially in recent years.The categories of BRM agents include antisera.monoclonal antibodies.cytokines.growU1 factors.fusion proteins, adoptive immunotherapy.somalic cell therapy, gene therapy, immunomodulating drugs and products derived from, or related to. the preceding types of products.At present, BRMs are in use or under clinical investigation for llie treatment of patients with a broad variety of conditions including: malignancies: infectious diseases of viral, bacterial and fungal origin; transplantation of bone marrow and solid organs: and immunodeficiency diseases.Thus, it is difficult to generalize about the clinical development of BRMS beyond saying that general principles of good clinical trial design, conduct and analysis should be applied.
Nevertheless, the property of being a BRM has important implications for all aspects of clinical development.Most notable are the selection of animal models , defining the study population, evaluating adverse reactions.dosing and defining end-points.Additionally, BRMs that are proteins, living cells and/or analogues of endogenous substances raise additional issues in clinical development as a result of lliese properties.

ANIMAL MODELS
Although testing in animal models is not actually a clinical trial.animal testing is so crucial to clinical development that it is worlliwhile to consider briefly some key issues in this context.BRMs can present many difficulties in animal testing, particularly as a result of species restriction and immunogenicity.A BRM may be species restricted as a result of absent or decreased affinity of llie homologous receptor (eg, a cyt.okine such as granulocyt.e-macrophage colony-simulating factor [GM-CSP] or interferon [wN]-y).absence or alteration of llie target antigen (eg, for a monoclonal antibody) and/or major histocompatibility complex restriction (eg.for a lymphocyte clone).Interpretation of long tenn studies of protein BRMs may be complicated by the formation of antibodies altering pharmacokinetics and/or pharmacodynamics.Cellular therapies may be subject to lack of activity due to species restriction and/or altered survival and function resulting from xenogeneic cellular and humoral immune responses.
There is relatively little information to be gained from studying species restricted BRMs in species in which they lack BRM activity.Such studies may yield some information regarding reactions due to impurities and reactions unrelated to llie intended mode of action.However, since many BRMS are quite pure, and since most species restricted BRMS act at low concentrations through high affinity receptors or ligands, the most 6A Y•DONOTCGn important adverse and desired reactions to these llierapies in humans are almost invariably related to specific binding.Such effects can only be stud ied in species in which binding and biological activity of the BRM occur.
Due to the difficulties in animal testing of BRMS.many are introduced into clinical trials with little relevant animal dala.The resultant paucity of information regarding beneficial and toxic effects, llierapeutic and toxic doses, and effects of various doses, routes, schedules and combinations on llie desired BRM effect requires lliat clinical studies proceed in a s low and cautious manner to ensure safety.Thus, the slarling dose will be low, escalation will be s low and obtaining answers to key pharmacological questions will be slow, costly, inefficient and sometimes dangerous.While such clinical difficulties may at times be unavoidable.their undesirability provides a strong impetus for testing BRMS in an appropriate animal model where possible.For species restricted BRMS, production and testing of the homologous animal product.while not part of llie classic toxicological testing, may provide the most valuable information about the likely pharmacological effects of llie human analogue.For many relatively species restricted products, responses in primate species may adequately reflect human responses.Additionally, animals engineered to be responsive to the BRM, eg, transgenic animals expressing the relevant human antigen or receptor, athymic nude mice or severe combined immunodeficiency disease mice bearing BRM responsive human immune cells may provide useful information, but only to the extent lliat the model reflects human conditions.
It should be noted that even BRMs that act across species lines may show marked interspecies variation in dose response profiles or nature of effect.The former may result from differences in the affinity of the receptors or from other differences in the sensitivity of the species tested in evoking the measured biological response.For these reasons, standard approach es used for extrapolation of doses to humans , eg, body surface area, weight and allometric scaling, may not be reliable.InfonnaUon about U1e relative dose response a nd b inding affinities of llie cells of the test species and of humans can be of great value in extrapolating dosage from the test species to humans.Such information should be generated before clinical testing whenever possible.

THE STUDY POPULATION
In traditional drug development, testing in normal volunteers can be an extremely valuable early source of information about the pharmacology of a drug.This type of testing is far less common in the development of BRMs, largely because the potential 1isk to humans of a new BRM is rarely well characterized before testing in humans.Even when good animal models are available and species resbiction is not a problem, some k ey PEl80NAL USE questions , notably the degree and U1e clinical implications of immunogenicity of a human protein in humans.cannot be well assessed in anin1als.Where an unknown but significant possibility of substantial harm to the subject exists.testing on normal volunteers with no potential for direct benefit is problematic.To keep risks and benefits in an appropriate balance, BRMs are often first.tested in humans with a disease thal might benefit from therapy, frequently an ultimately fatal disease.
The target population for most BRMs includes immunocompromised patients, such as patients with advanced malignancies or immunodeficiency disorders, transplant recipients, recipients of myelotoxic therapies.neonates and the elderly.In studying BRMS in immunosuppressed patients it is important to bear in mind thal the effects of many BRMS.both desired and undesired, may be mediated by cells of the immune system.Patients in whom the target cells of the BRM are either absent or irreversibly hypofunclional may not respond to a BRM as would other individuals.ln general, the optimal target population for early trials of a BRM therapy consists of patients who have a serious or terminal disease that might benefit from lhe therapy, but in whom physiological functions , particularly drug metabolic and excretory functions, immunological functions and the ability to tolerate drug toxicities.are relatively intact.

ADVERSE REACTIONS
BRMS are capable of inducing a broad range of adverse reactions.Nevertheless, a number of generalizations can be made regarding adverse reactions to BRMS. and they should be born in mind in planning clinical development.
Firsl, certain patterns of toxicity appear lo be particularly common in BRM therapy.Central nervous system toxicities include agitation, seizure.lethargy, coma with interleukin (IL) -2 and mood changes with IFN -a.Vascular leak leading to effusions and edema have been observed with IL-2.GM-CSF and tumour necrosis factor (TNF)-a.Although the endocrinological effects of many cytokines have yet to be fully investigated , IL-2 therapy has been reported lo be associated with elevated levels of corticotropin.cortisol, prolaclin, growth hormone and beta-endorphin and decreased levels of thyroid hormone, while IFN-a therapy has been reported to be associated with lowered serum estradiol and progesterone levels (as well as menstrual irregularities, decreased fertility and increased spontaneous abortions in primates).Perhaps these patterns are not surprising in thal there are increasingly recognized close interactions between the immune system and the neurological and endocrine systems, and because regulation of vascular permeability is critical to regulation of immunological and inflan1matory responses .Other common toxicities of immunomodulating BRMs include fever and 'flu -like symptoms observed with a variety of cyt.okines including IFN-a , IFN-~.IL-2, TNF-a and GM-CSF.The similarity of toxicities observed with a variety of immunoregulatory cytokines and other BRMS should not be surprising in that cyt.okines are involved in a complex regulatory network in which there is considerable functional overlap; lhe administration of one cytokine oflen induces the production of several others cyt.okines.
Second, many BRM therapies have significant theoretical possibility of exacerbating disease.Exacerbation may occur through several mechanisms.Host defences are not only responsible for control of malignancies and infections.bul they may also cause pathology.Since the physiological regulation of hosl defences is poorly understood , attempts to improve natural defences, regardless of the effector mechanism targeted, may adversely affect disease control and/or may contribute lo tissue destruction or dysfunction associated with the disease.Thus, preliminary data in sepsis trials suggest some anti-inflammatory BRMS may be associated with higher mortality rates when administered lo some patients with sepsis.In malignancies in which the tumour cells may bear receptors for cytokines and growth factors employed in or induced by BRM therapy, (eg, many hematological malignancies), the possibility thal the BRM will promote growth of the tumour musl be considered.
Third, BRM therapy may lead to the production of antibodies.BRMs that are proteins are oflen immunogenic.Even those proteins that are very closely related to endogenous substances may be immunogenic due to the unusual dose, schedule and route of presentation to the immune system.Antibody formation is particularly likely after repeated, prolonged and intem1illent administration, and should be assessed for all protein BRMs.When antibodies are detected they should be tested for their ability to neutralize the BRM and, where relevant, its natural analogue.Furthermore.it is important to study patients developing antibody reactions for the following potential consequences: Joss of response (therapeutic or adverse) to the BRM: change in pharmacokinetics of the BRM; immune complex disease; and loss of activity of the endogenous natural analogue of the BRM.Information on the duration of the antibody response and.when they can be obtained safely, data regarding the effects of read ministration can be of greal value.In addition to the immunogenicily of proteins.the immunogenicity of cellular therapies.whether modified aulologous.allogeneic or xenogeneic cells, is a major concern under active study.

DOSING
As noted above , perfom1ing studies in a relevant animal model.and delerrnining the quantitative relationship of dose response effects on animal and human cells, can be of great value in determining a safe starting dose and escalation rate thal will allow early sludy of biologically active doses .Determining the optimal dose (as well as route and schedule) of a BRM to carry forward into further phase 2 and 3 trials can be difficult.Typically, chemotherapeutic agents for the treatment of malignancies or serious infections are tested and ultimately used at maximally tolerated doses.In the area of BRM therapies.it has long been argued that more is not always better and that the optimal dose of a BRM is the lowest dose causing a maximal modification of targeted biological response.
Determining an optimal biological response modifying dose requires having a relevant., valid and reproducible assay of the biological response under consideration or a closely related biological effect.Functional assays of cells are notoriously difficult to make standard and reproducible, particularly in multicentre studies where multiple laboratories or transport of living cells must be used .Where possible, biological effects should be measured by cell counts, serum proteins or chemistries, or chemical or molecular rather than functional studies on cells.Considerable efforts should go into standardizing and ensuring reproducibility of these assays.
Additionally, is to ensure that any biological effect used to guide dosing or other key aspects of drug development be relevant.Although it is desirable that the biological effect assayed be the response that is hypothesized to cause clinical improvement, this is not always feasible .In such cases, one should ensure that the monitored biological effect depends on the same receptor and, preferably, the same effector cell, as the targeted response.

END-POINTS
As noted in the section on dosing, measurement of the biological effects of a BRM can provide an end -point of great value in the clinical testing of a BRM.Such effects, if reliably measured and related to clinical outcome, can be used to optimize dose, route, schedules, combination thera py and target population.Of course.determinations of efficacy of a BRM must be based on end-points that reflect meaningful changes in disease course or symptoms.Beginning with the early trials of 8A a BRM, one should put considerable efforts into correlating measured biological effects with these clinically significant events.Evidence of such correlations can help select which biological effect to optimize in future drug development and may allow use of early biological effect measurements to identify those patients most likely to benefit from ongoing therapy.When the biological effect can be measured reproducibly and the correlation with clinical events is strong and consistent, it may b e valid to use the biological response measurement as a surrogate for clinical efficacy, thus greatly facilitating the clinical investigation of the BRM and closely related therapies.

SUMMARY
The clinical investigation of BRMs has been underway for many years and has accelerated recently in parallel with advances in biotechnology.The following generalizations can be made regarding the clinical development of BRM therapies: • Animal models for testing BRMs are often difficult and costly to develop and validate.
Neve rtheless, such efforts can be extremely worthwhile in facilitating and expediting clinical testing.
• While the initial human testing of BRMs is often most appropriately done in patient populations, it is generally d esirable that these patients have relatively intact physiological functions.
• Certain classes of BRMS induce characteristic types of adverse reactions that should be studied carefully.
• A reliable, reproducible measure of the intended biological response modification or a related biological effect can be of great value in drug development, particularly in selecting an optimal dose and regimen to evaluate for clinical efficacy.
• Studies of the relationship between measurable biological effects induced by a BRM and clinical benefit should be actively pursued throughout clinical testing.
CAN J INFECT DIS VOL 5 SUPPL A FEBRUARY 1994 LY•DO COPY Clinical development of BRMs