0% found this document useful (0 votes)

108 views22 pages

Industry Perspective On First-In-Human and Clinical Pharmacology Strategies To Support Clinical Development of T-Cell Engaging Bispecific Antibodies For Cancer Therapy

This white paper discusses the clinical development of T-cell engaging bispecific antibodies (TCEs) for cancer therapy, highlighting their promise in treating hematological cancers and the challenges faced in solid tumors. It emphasizes the complexities of selecting first-in-human (FIH) starting doses and the importance of pharmacological strategies, including the use of in vitro assays to inform dosing decisions. The paper also reviews recommendations for optimizing clinical pharmacology considerations to enhance the efficacy and safety of TCEs in early clinical development.

Uploaded by

hiusifhd123

We take content rights seriously. If you suspect this is your content, claim it here.

Available Formats

Download as PDF, TXT or read online on Scribd

0% found this document useful (0 votes)

108 views22 pages

Industry Perspective On First-In-Human and Clinical Pharmacology Strategies To Support Clinical Development of T-Cell Engaging Bispecific Antibodies For Cancer Therapy

Uploaded by

hiusifhd123

We take content rights seriously. If you suspect this is your content, claim it here.

Available Formats

Download as PDF, TXT or read online on Scribd

You are on page 1/ 22

WHITE PAPER

Industry Perspective on First-in-Human and

Clinical Pharmacology Strategies to Support
Clinical Development of T-Cell Engaging
Bispecific Antibodies for Cancer Therapy
Prathap Nagaraja Shastri1,*, Nirav Shah1, Martin Lechmann2, Hardik Mody3, Marc W. Retter4, Min Zhu5,
Tommy Li6 , Jun Wang7, Naveed Shaik8, Xirong Zheng9, Meric Ovacik10, Fei Hua11 , Vibha Jawa9,
Christophe Boetsch12 , Yanguang Cao13 , John Burke11, Kaushik Datta14 , Kapil Gadkar10,
Vijay Upreti15 and Alison Betts16

T-cell-engaging bispecific antibodies (TCEs) that target tumor antigens and T cells have shown great promise in
treating cancer, particularly in hematological indications. The clinical development of TCEs often involves a lengthy
first-in-human (FIH) trial with many dose-escalation cohorts leading up to an early proof of concept (POC), enabling
either a no-go decision or dose selection for further clinical development. Multiple factors related to the target,
product, disease, and patient population influence the efficacy and safety of TCEs. The intricate mechanism of
action limits the translatability of preclinical models to the clinic, thereby posing challenges to streamline clinical
development. In addition, unlike traditional chemotherapy, the top dose and recommended phase II doses (RP2Ds)
for TCEs in the clinic are often not guided by the maximum tolerated dose (MTD), but rather based on the integrated
dose–response assessment of the benefit/risk profile. These uncertainties pose complex challenges for translational
and clinical pharmacologists (PK/PD scientists), as well as clinicians, to design an efficient clinical study that guides
development. To that end, experts in the field, under the umbrella of the American Association of Pharmaceutical
Scientists, have reviewed learnings from published literature and currently marketed products to share perspectives
on the FIH and clinical pharmacology strategies to support early clinical development of TCEs.

T-cell-based immunotherapies, including immune-checkpoint in- at high concentrations (doses) of TCEs driving a reduction in cy-
hibitors (ICIs), chimeric antigen receptor T cells (CAR-Ts), and totoxic response.
T-cell engagers (TCEs) have demonstrated great promise in cancer Since the proposed concept of bispecific antibodies (bsAbs)
treatment.1 CD3-based bispecific TCEs bind to tumor-associated in the 1960s and the subsequent development of hybridoma and
antigens (TAAs) on tumor cells with one arm and to CD3 on T knobs-into-holes technologies, TCEs have shown significant ad-
cells with another arm. The resulting formation of immunological vancement in clinical development over the past few decades.
synapses leads to T-cell activation and the release of inflammatory Blinatumomab, which targets CD19 and CD3, was approved by
cytokines and cytolytic molecules that induce the killing of tumor the US Food and Drug Administration (FDA) in 2014 for the
cells.2 Efficacy and on-target toxicity are driven by the formation treatment of Philadelphia chromosome-negative relapsed or re-
of a trimolecular complex (hereafter trimer) between the TCE, fractory B-cell precursor acute lymphoblastic leukemia (ALL).
T cell, and tumor cell at the immune synapse. The formation of Recently, multiple TCEs, including teclistamab (BCMA × CD3),
the trimer leads to a complex mechanism of action, including a monsunetuzumab (CD20 × CD3), epcoritamab (CD20 × CD3),
bell-shaped dose–response, with a reduction of trimer formation glofitamab (CD20 × CD3), talquetamab (GPRC5D × CD3), and

1
Clinical Pharmacology and Pharmacometrics, Johnson and Johnson Innovative Medicine, Spring House, Pennsylvania, USA; 2Roche Pharma Research
and Early Development, Pharmaceutical Sciences, Roche Innovation Center Munich, Penzberg, Germany; 3Clinical Pharmacology, Genentech, South San
Francisco, California, USA; 4Preclinical PK/PD, Regeneron Pharmaceuticals, Tarrytown, New York, USA; 5Clinical Pharmacology, Regeneron Pharmaceuticals,
Tarrytown, New York, USA; 6Clinical Pharmacology, Genmab, Plainsboro, New Jersey, USA; 7Biotherapeutics Discovery Research, Boehringer Ingelheim
Pharmaceuticals, Inc, Ridgefield, Connecticut, USA; 8Clinical Pharmacology, Pfizer Oncology Development, San Diego, California, USA; 9Clinical
Pharmacology, Pharmacometrics and Bioanalysis, Bristol Myers Squibb, Princeton, New Jersey, USA; 10Preclinical PK/PD, Genentech, South San Francisco,
California, USA; 11Certara Predictive Technology, Certara, Concord, Massachusetts, USA; 12Roche Pharma Research and Early Development, Pharmaceutical
Sciences, Roche Innovation Center Basel, Basel, Switzerland; 13Division of Pharmacotherapy and Experimental Therapeutics, Eshelman School of Pharmacy,
University of North Carolina at Chapel Hill, Chapel Hill, North Carolina, USA; 14Nonclinical Safety, Bristol Myers Squibb, Princeton, New Jersey, USA; 15Clinical
Pharmacology, Modeling & Simulation, Amgen, South San Francisco, California, USA; 16Preclinical & Translational Sciences, Takeda Pharmaceutical
Company Limited, Cambridge, Massachusetts, USA. *Correspondence: Prathap Nagaraja Shastri ([email protected])
Received May 27, 2024; accepted August 25, 2024. doi:10.1002/cpt.3439

CLINICAL PHARMACOLOGY & THERAPEUTICS | VOLUME 0 NUMBER 0 | Month 2024 1

elranatamab (BCMA × CD3) have received regulatory approvals overly conservative starting dose and the time to reach a clinically
for the treatment of lymphoma or multiple myeloma. The latest efficacious dose or maximum tolerated dose (MTD) needs to be
addition to the list is tarlatamab (DLL3 × CD3) which received considered. For TCEs, unlike conventional cytotoxic chemother-
accelerated approval by the FDA for extensive stage small cell lung apy drugs, it’s anticipated that the optimal balance between safety
cancer. and efficacy may be achieved at doses below the MTD. However,
Despite the success of TCEs in treating hematological indica- given the mechanism of action (MoA), the use of TCEs in differ-
tions, the development of TCEs in treating solid tumors is still in ent disease settings, as well as the potential combination therapies
its infancy, due to constraining factors such as limitations in the with other agents, complexities continue to exist in selecting the
biodistribution of TCEs to the site of action, tumor heterogeneity, appropriate doses in expansion cohorts, as well as identifying the
immunosuppressive tumor microenvironment (TME), and lim- recommended phase II dose (RP2D) for further exploration in
ited trafficking of T cells into solid tumors. The sparse infiltration pivotal studies.
of T cells in certain types of solid tumors, along with the impaired To address the challenges associated with TCEs discussed above,
tumor-infiltrating lymphocyte (TCL) quality, prevent the forma- a working group was formed under the umbrella of the American
tion of effective immunological synapses and thus potentially limit Association of Pharmaceutical Scientists to review learnings from
the clinical efficacy of TCEs. Despite these challenges, preliminary published literature, currently marketed products as well as au-
efficacy has been reported for tarlatamab (DLL3 × CD3)3 and xal- thors’ own experience. While there are other T-cell redirecting
uritamig (STEAP1 × CD3)4 in small cell lung cancer and prostate antibodies such as those targeting γδ T cells, or those binding to
cancer, respectively. 4-1BB, CD28 co-stimulatory domains, the focus of this white
Given the immune agonistic properties of TCEs, minimal an- paper is to share perspectives on the FIH starting dose and clinical
ticipated biological effect level (MABEL), which is predominantly pharmacology strategies including the selection of RP2D to sup-
based on in vitro pharmacological activity assays, has been widely port early clinical development of CD3-based TCEs (Figure 1).
used by sponsors to select the first-in-human (FIH) starting doses We summarized learnings and recommendations in three sections:
of TCEs. While this approach has generally led to starting doses FIH starting dose selection, clinical pharmacology consideration
with manageable safety profiles, a balance between selecting an and bioanalytical considerations.

Figure 1 Schematic representation of clinical pharmacology considerations toward FIH and clinical development of T-cell engaging bispecific
antibodies.

2 VOLUME 0 NUMBER 0 | Month 2024 | www.cpt-journal.com

FIH STARTING DOSE CONSIDERATIONS effector cells or tumor resident T cells have sourcing, viability,
Traditional MABEL approach and variability issues; however, if feasible to use, these could pro-
To select a safe FIH starting dose for TCE molecules, tradition- vide a pertinent simulation of patient tumor microenvironment
ally a MABEL-based approach is recommended due to their im- in an in vitro setting.
mune agonist properties. As per the recommendation, a starting The impact of choosing the most relevant target and effec-
dose is selected based on matching simulated human exposure to tor source can be exemplified by the case study of teclistamab
the EC20 or EC30 from a relevant in vitro pharmacology assay. (BCMA × CD3) for which a FIH starting dose of 0.3 μg/kg was
This pharmacology-informed exposure-based MABEL approach chosen, based on EC20 of cytotoxicity from MM.1R cell line in-
has been shown to be safer than that based on receptor occupancy cubated with purified healthy donor T -cells at a 5:1 E:T ratio.
(RO), or animal toxicology (NOAEL/HNSTD) studies in a ret- Recently approved in 2022, the dosing regimen of teclistamab
rospective analysis of 17 bispecific constructs published by Saber starts at a step-up dose of 60 μg/kg,6 which is 200× higher than the
et al.5 Although this approach has resulted in safe FIH starting FIH starting dose. A retrospective look at the in vitro data gener-
doses in patients, this strategy often comes with challenges. The ated for this molecule suggests that using PDCs could have led to a
experimental conditions selected for the in vitro assays can in- starting dose 10 × >0.3 μg/kg even at EC20 potency.6
fluence the MABEL-based starting dose making it difficult to Another key consideration is the choice of the effective concen-
achieve a “one-size-fits-a ll” approach for TCEs. This is because tration (ECx) as the potency indicator of the MABEL readout.
the efficacy of TCEs depends upon trimer formation, which is Saber et al. recommended EC20 or EC30 of the most sensitive assay
impacted by variables including receptor expression, effector as the threshold to determine MABEL. However, it was concluded
(E): target (T) cell ratio, and the binding affinity of the TCE to that EC50 was acceptable for MABEL for all but one of the re-
CD3/TAA, as well as TCE concentration. Differences in these viewed TCEs. The one TCE that did not provide an acceptable
variables across assays will impact the responses observed, and the starting dose at EC50 was a tetravalent construct, highlighting the
resulting translation to the clinic. Bell-shaped dose–response re- influence of heterogeneity in the format of TCE potentially con-
lationships are frequently observed in preclinical in vitro and in tributing to decreased tolerability in patients. For the case studies
vivo studies compared with clinical settings. This discrepancy can highlighted in Table 2, since dose escalation was safely managed
be attributed to factors such as higher achievable doses, controlled in the clinic, it is implied that EC50 would have led to a safe higher
experimental conditions (e.g., consistent effector-to-target ratios, starting dose. An EC50 potency choice for MABEL concentration
healthy T-cell populations, standardized tumor burdens), and the may be justifiable if the target expression pattern on non-tumor tis-
relatively static nature of preclinical models compared with the sues is well characterized. Given the variability between patients,
complex dynamics of human disease. As such, it is important to it is also recommended that EC50 values from a range of in vitro
consider the physiological relevance of the experimental condi- assays be considered.
tions to obtain a starting dose that is not only safe but also closer In a typical in vitro functional assay for TCEs, the effector and
to pharmacological effects. This is particularly important as FIH target cells are combined in a particular proportion referred to as
trials enroll cancer patients who are suffering from advanced the E:T ratio. The choice of relevant E:T ratio is critical as a high
forms of cancer. Table 1 provides a comprehensive perspective, proportion of effector cells relative to target cells (3:1, 5:1, 10:1)
caveats, and general recommendations for optimization of the extends to overestimate the potency of the molecule7 ultimately lead-
perimental conditions to be used for MABEL-based starting dose ing to a lower starting dose. E:T ratios are not only unique to every
calculation. indication but also depend on each patient’s immune status and
One of the key factors for the optimization of experiments target burden. Moreover, the standard of care treatment for can-
for MABEL determination is the choice of target and effector cer patients is immune ablating, which, compounded by increasing
cell sources. Patient-derived cells/tumors (PDCs) are biologi- tumor burden, tends to result in a much lower E:T ratio. The E:T
cally more relevant than cell lines in terms of target expression ratios are further attenuated for cold solid tumors which may be
level and tumor heterogeneity and sensitivity to T-cell killing. less accessible to the immune cells. Table 1 further describes the
Additionally, using cell lines with target densities that are not key considerations for E:T ratios including differences between
clinically relevant could result in in vitro potencies that may not solid and hematologic tumors. A case study from Amgen evaluat-
be relevant to the patient population. However, sourcing issues ing a TCE for gastric cancer emphasizes the impact of optimal E:T
and sample variability can pose challenges with PDCs, especially ratios and their influence on the starting dose.8 Cytotoxicity assays
solid tumor-derived cells, making cell lines technically more fea- performed at E:T ratios of 10:1 resulted in a 10 × lower starting
sible. In the absence of PDCs, careful consideration needs to be dose than physiologically more relevant E:T ratio of 1:5. Overall,
given in choosing representative cell lines with physiologically lower E:T ratios such as 1:1 for hematological malignancies and
relevant target-level expression. Similarly for effector cell source, 1:5 or 1:10 for solid tumors could potentially enable a more clin-
the use of peripheral blood mononuclear cells (PBMCs) is pre- ically relevant MABEL selection as long as low E:T ratios do not
ferred over purified and/or activated pan T cells. The presence compromise the observation of functional signals in the in vitro
of a global population of immune cells in PBMCs makes them assays.9
a very appropriate source of cytotoxic T cells. In cases where sol- Choosing the most relevant assay readout also influences the
uble target engagement is expected, whole blood may be more choice of starting dose. Typically, T-cell activation (CD69+ and
informative although challenging to work with. Patient-derived CD25+), cytotoxicity, and cytokine release are evaluated, and the

CLINICAL PHARMACOLOGY & THERAPEUTICS | VOLUME 0 NUMBER 0 | Month 2024 3

Table 1 Recommendations on in vitro pharmacological assay systems to determine clinically relevant MABEL for
CD3-bispecific antibodies
Factors impacting
MABEL Question Perspective/recommendation Caveats/Considerations
Target source What is the ideal Primary patient-derived cells/tumors Procurement/sourcing and viability of PDCs is a
and expression source of target – (PDCs) are physiologically more relevant key limitation in general but more so for solid tu-
levels primary cells or cell and preferred over cell lines mors. PDCs could present high variability requir-
lines? In the absence of PDCs, cell lines with ing a statistically relevant approach to account
patient-relevant target expression levels for variability based on totality of data47
can be an alternative source of target Cell lines are more convenient to use but are
cells typically more sensitive to TCEs and need to be
evaluated at physiological expression levels
Effector source What is the ideal PBMCs from patients are an appropriate It is important to capture donor-to-donor
source of effector effector cell source in general. variability. Effector cells from patients could
cells – whole blood, However, in cases where soluble target be less effective than effectors from a heathy
PBMCs, or purified/ engagement is expected, patient whole donor, and may also lead to a relevant measure
activated T cells? blood matrix if available could be more of expected clinical activity (e.g., soluble BCMA
informative concentrations impacting bispecific activity)
E:T ratios What is the ideal E:T It is generally recommended to assess In general, E:T ratios for hematologic indication
ratio? E:T ratios in patient samples for the could be higher than E:T ratios for solid tumors
What is the impact of intended indication and use the relevant but are likely dependent on target burden and
E:T ratios on in vitro E:T ratio for MABEL experiment quality of effector cells. It is however possible to
functional assays In the absence of primary samples, have a highly variable range within an indication
readouts? published reports could be used for such as reported in Krupka et al.,48 ranging from
guidance. In general, E:T ratios within an 1:3–1:80 in samples from the AML patients
experimentally feasible range of 5:1 to E:T ratios could have a directly proportional
1:5 is accepted for the MABEL experi- impact on both EC50 and Emax when it comes to
mental setup cytotoxicity.8 In such cases, potency estimates
If generating in vitro data at relevant E:T from T-cell activation alone can support MABEL
ratios are not experimentally feasible estimate
due to the lack of dose–response at Additionally, target expression in off-tumor areas
lower E:T ratios, one could explore QSP should be considered while defining E:T ratios
models to estimate the dose–response
at relevant E:T ratios
Incubation time How do we pick time It is crucial to determine timepoints at Characterizing potency at various timepoints can
points? which the pharmacological readouts typi- provide insights into kinetics and dynamic changes
cally reach Emax (i.e., Tmax – time to max. leading to the pharmacological event as illustrated
effect). The relevant timepoint can vary by Van de Vyver et al.23 Perhaps, such assays are
based on the molecular design, target, more useful to determine an ideal experimental
and the assay setup setup for the FIH enabling MABEL assay
An experimental setup that extends Additionally, cytokine readouts often tend to
much longer than the Tmax for in vitro increase with time give the static nature of the
activity may lead to an artificially inflated in vitro setup. A 700 to 4,000-fold difference is
potency estimates affecting the starting observed in vitro assays,49 and the in vitro – in
dose vivo correlation is poorly understood. Therefore,
for cytokines, early timepoints such as 24 or
48 h are likely mostly relevant
Readouts What are the relevant Of cytokines, cytotoxicity, and T-cell While cytotoxicity is the most relevant readout to
readouts to derive activation, the most sensitive readout is determine an effective dose, at lower E:T ratios,
MABEL? generally picked while defining MABEL it is challenging to get a suitable dose–response
How to determine if in However, in the case of choosing cy- to determine a MABEL based on cytotoxicity
vitro cytokine elevation tokines, the physiological significance of assay
is the most sensitive the elevated cytokine should be taken For T-cell activation, both CD25 and CD69
readout – should it into consideration. Given the dynamic activation markers have been used. Being an
based on the most nature of cytokines in terms of their early marker, using CD69 T-cell activation would
relevant cytokine or function and pharmacological effects, be relevant to determine early activity after the
set of cytokines? we do not recommend singling out any starting dose
cytokines Exposure levels at which consistent elevation
IL-2, IL-6, IL-10, TNFa, and IFNg are across multiple relevant cytokines (e.g., IL-6,
commonly assessed. IL6 and other pro- TNFa, and IFNg) are observed can be used as
inflammatory cytokines (e.g., IL-1b) are a MABEL threshold when cytokine readouts are
tied to CRS mechanisms and are known most sensitive. Additional cytokines may be
to be clinically relevant, as evident from relevant, such as GM-CSF. However, it is unclear
the use of antagonists of IL-6 and IL-1b how its production in vitro may translate to the
in the clinic to mitigate CRS in vivo setting

(Continued)

4 VOLUME 0 NUMBER 0 | Month 2024 | www.cpt-journal.com

Table 1 (Continued)
Factors impacting
MABEL Question Perspective/recommendation Caveats/Considerations
5
Potency/ECx What EC value In 2017, Saber et al. recommended Assay setups have a higher impact on the final
used to define is generally EC20 or EC30-based MABEL. However, MABEL estimate than the ECx value used.
MABEL recommended to FDA’s recent meta analysis as presented Choosing EC50 over EC30 will only yield a
define MABEL? at the American College of Clinical difference of ~2–4-fold which is generally within
Pharmacology indicated that EC50 the experimental variability. For novel targets
based MABEL was generally accept- with known expression in off-tumor areas, a
able when a pharmacological relevant conservative approach (e.g., EC20 or EC30) may
assay systems was employed. In ad- be warranted
dition, across the case studies listed
in Table 2, EC50 based starting doses
would be a safe dose for FIH
Therefore, we recommend defining
MABEL at EC50 using pharmacologically
relevant MABEL assay setup

most sensitive readout is often selected to determine the MABEL of in vitro or in vivo data. These approaches still must adhere to
concentration for the starting dose. If T-cell activation or cytotox- the starting dose principle of establishing a minimal biological
icity are the most sensitive readouts, the choice is more straight- effect level. Such evolving approaches can be categorized as “Next-
forward. However, there are a few caveats that need to be taken generation MABEL” and are discussed in this section.
into consideration while picking cytokine readout as MABEL As we discussed translational limitations of the traditional
even when it may be the most sensitive functional readout. The MABEL approach in ‘Traditional MABEL approach’, starting
in vitro elevation and sensitivity of cytokines can be influenced by dose selection based on in vivo data, specifically cytokine release
continuous “bathing” of the target and effector cells in accumu- in NHPs can be an alternative approach. A successful application
lated cytokines. This potentially could be the reason why in the of this approach is demonstrated by the case of glofitamab, where
case of glofitamab (CD20 × CD3), the in vitro IL-6 release was a PK/PD model was applied to the observed cytokine release in
found to be 70-fold more potent than observed in non-human NHPs to determine the FIH starting dose (Table 4). The spon-
primates (NHPs).10 Cytokine release can be more sensitive in in- sors captured the in vivo IL-6 release in NHPs and translated it to
dividual donors but is also highly variable between donors. Out humans to estimate the starting dose at which the IL-6 levels in hu-
of a variety of cytokines evaluated, an in vitro increase in one of mans stayed below a predefined threshold of 600 pg/mL. The IL-6
the cytokines may not necessarily mean a risk of cytokine release threshold was clinically predefined and estimated to be well toler-
syndrome (CRS) in patients. For talquetamab (GPRC5d × CD3), ated.10 The FIH starting dose was further adjusted by considering
IL-10 was the most sensitive readout, with an EC20 of 0.032 nM. the markedly higher in vitro potency for glofitamab in humans vs.
However, since none of the other cytokine levels were as sensitive NHPs and a 10-fold safety margin was included to compensate for
and there was high donor variability, the sponsor chose a starting potential cytokine release from tumor. By doing so, the sponsor
dose based on EC20 of CD25 marker for T-cell activation which was able to justify a starting dose ~30 × higher than the in vitro
led to ~2.5× higher dose.11 It could be more appropriate to choose cytotoxicity EC50. However, a key requirement for this approach
cytokine release as a readout when there is a collective increase in is the similarity between NHPs and humans in terms of expres-
some or all the key cytokines implicated in the initiation of cyto- sion and distribution of TAA, and the cross-reactivity of both TCE
kine storm such as TNF-α, IFN-γ, GM-CSF, IL-1, and IL-6 (4–8). arms. Even then, the target burden in patients is expected to be
For odronextamab (CD20 × CD3), IL-6, INF-γ, and TNF-α lev- higher than in NHPs and should be considered for translation of
els were measured in vitro and reported to have EC50 values in the in vivo cytokine release. In addition, prior clinical experience and
range of 0.3–0.5 nM.12 In retrospect, exposures corresponding to characterization of a target could benefit starting dose justification
these EC50 values were well tolerated in the clinic along with nota- for subsequent next-generation molecules toward the same target.
ble cytokine release in NHL patients. In summary, cytokine release This is demonstrated for CEACAM5-TCB (Table 2), where reg-
can serve as a MABEL readout if relevant cytokines demonstrate ulators accepted a starting dose that was justified from the clini-
significant increases in in vitro experiments. cal experience with another CEA targeting TCE-cibisatamab.
However, diligence is still required by considering the in vitro
Next-generation MABEL approach potency differences, off-tumor toxicities observed on healthy cells
As discussed in the previous section, certain challenges with the (e.g., whole blood cell assay and primary intestinal cells co-cultured
traditional MABEL approach can be overcome by well-designed with PBMCs) as well as pharmacokinetic properties to adjust the
in vitro assays to derive a safe and relevant starting dose. However, starting dose as was done for CEACAM5-TCB (unpublished).
opportunities exist to further advance the selection of FIH doses A QSP modeling-based approach is a valuable tool to over-
by leveraging pharmacokinetic/pharmacodynamic (PK/PD) and come some of the translational limitations of the preclinical
quantitative systems pharmacology (QSP) models for translation models.13 QSP models are increasingly being used to support

CLINICAL PHARMACOLOGY & THERAPEUTICS | VOLUME 0 NUMBER 0 | Month 2024 5

Table 2 Review of experimental conditions and approaches used for FIH starting dose selection for CD3-bispecific TCEs
FIH Starting dose Effector MABEL
starting calculation Target cells/ readout (ECx;
dose approach source source timepoint) E:T Lessons learnt
Teclistamab IV – 0.3 Traditional MABEL Cell Purified T Cytotoxicity 5:1 Using PDCs could have led to a
(BCMA)6 μg/kg, lines cells from (EC20; 48 h) higher starting dose
Q2W HD SD derived based on primary
samples and EC50 is close to the
current step-up dose
Odronextamab IV – 30 μg Traditional MABEL Cell WB from Cytotoxicity 5:1–10:1 EC50 values for IL-6, TNF-α,
(CD20)11 line HD, PBMCs (EC50) and IFN-γ derived from In vitro
WB cytokine assay EC50 values
correlated with exposure levels
associated with cytokine release
in patients and could have led
to a higher and relevant starting
dose. Use of IL-6, TNF-α , and IFN-γ
as representative cytokines
Talquetamab IV – 0.5 Traditional MABEL Cell WB from T-cell 5:1 IL-10 EC20 was the most sensitive
(GPRC5d)10 μg/kg, line HD activation readout. However, this was not
Q2W (EC20; 48 h) chosen as MABEL concentration
due to the pleotropic functions
of cytokines and high inter-donor
variability in the cytokine EC
values. Most relevant readout >
most sensitive readout
Glofitamab IV – 5 μg, Next generation NA NA Cytokine NA A PKPD model was used to scale
(CD20)9 Q3W (NHP PK/ release IL6 dose/exposure–response
PD- based) In vivo (NHP) from NHP to human. IL6 release
was found to be 70-fold more
potent in human than in NHPs. A
safety factor of 10 was added to
the SD calculation to compensate
for potential cytokine release from
tumor. SD derived using in vivo
IL6 release in NHPs is 35x higher
than in vitro MABEL
p- Cadherin13 IV – Traditional MABEL Cell PBMCs Cytotoxicity 5:1 QSP model was able to
1.5 ng/ line (EC20) simultaneously describe kinetics
kg, QW Next generation Cell PBMCs Cytotoxicity 5:1, 2:1, 1:1 of tumor cells and T cells at
(QSP-based) line (EC20, trimer (simultaneously various E:T ratios and estimated
formation) fit) the in vitro synapse- based
MABEL. Dose estimations using
both traditional MABEL and
model-based approaches yielded
similar starting dose. However, in
retrospect, QSP model would have
enabled a much higher starting
dose if the dose was selected
based upon the trimer formation
following the first dose rather than
steady state
AMG bispecific IV, QW Traditional MABEL Cell NA T-cell 1:5 At a higher E:T ratio of 10:1, the
gastric cancer8 line activation resulting starting dose was 10x
(CD69) lower. E:T optimization in the
(EC50) MABEL assays could help chose
clinically relevant starting dose
Cibisatamab 52 μg Traditional MABEL Cell PBMCs (30 Cytotoxicity 10:1 Choosing EC50 for MABEL would
(CEA)50 line donors) (EC20, 24 h) have led to a higher starting
dose (3.4 times higher). Deriving
MABEL based on WB assay or
primary cells would have led to
higher MABEL concentration

(Continued)

6 VOLUME 0 NUMBER 0 | Month 2024 | www.cpt-journal.com

Table 2 (Continued)
FIH Starting dose Effector MABEL
starting calculation Target cells/ readout (ECx;
dose approach source source timepoint) E:T Lessons learnt
CEACAM5-TCB 65 μg Defined Cell PBMCs (30 Cytotoxicity 10:1 Applying clinical learnings from a
(CEA) acceptable line donors) (EC20, 48 h) target for FIH SD determination
(unpublished) starting dose of a next-generation molecule for
for Cibisatamab the same target. In vitro potency
and account for differences may need to be taken
20-times higher into account
potency compared
with Cibisatamab

decision-making across the drug discovery and development pro- to a MAGE-A4 peptide displayed on HLA-A*02:01 on cancer
cess, from molecule design to clinical dose optimization (Table 4). cells and monovalently to CD3 on T cells.15
The advantage of QSP modeling lies in its ability to capture the Betts et al. presented two case studies for TCEs (for undisclosed
dynamics of trimer formation which is the key driver for efficacy targets) where the QSP model helped to justify a higher starting
and toxicity. The QSP model often uses the same in vitro data as dose based on model-estimated trimer formation compared with
those in the traditional MABEL approach. However, instead of traditional MABEL approaches in both hematological and solid
relying on exposure comparisons between in vitro experiments tumors when the same EC criteria was used for both QSP model-
and humans to establish the starting dose, the QSP model-based based and traditional MABEL approaches (data not shown).16
approach estimates trimer formation at different TCE concentra- Most current QSP models for TCEs employ trimer formation to
tions (doses) and experimental conditions (E:T or receptor ex- efficacy and toxicity. However, instead of the intermolecular tri-
pression), which can then be linked to either tumor cell killing or mers, the intercellular immunological synapse serves as a closer
cytokine release. By using model-estimated trimer formation as a driver of pharmacology. The formation of these synapses between
metric, the QSP model effectively bridges the gap between human tumor cells and T cells involves a complex cascade of molecular
and in vitro data for MABEL determination. The model incorpo- and cellular interactions that go beyond the antigen-binding mo-
rates multiple factors discussed previously, such as the local density lecular events. QSP models continue to evolve and become more
of effector and target cells (result of different E:T ratios), and rate sophisticated (e.g., pooled receptor approach vs. synapse-centered
of effect/target cell encounter, TAA expression levels on both nor- approach)17 in terms of capturing TCE MoA and thus can play an
mal and tumor cells, target turnover rate, binding kinetics, synapse important role in further optimizing the selection of the FIH start-
formation, and subsequent break down of synapse. The advantage ing dose. Especially for disease indications where the physiologi-
of using the QSP model in translation and to inform FIH dose se- cal E:T ratios are low or relevant tumor cell lines are not available,
lection is that the model parameters can be varied among in vitro, MABEL assays can be conducted at technically feasible E:T ratios
preclinical experiments and cancer patients to reflect the known and/or using multiple tumor cell lines. These data can in turn be
differences in biology such as E:T ratios and TAA expression lev- used to extrapolate the trimer kinetics at physiologically relevant
els. Subsequently, it estimates different concentration vs. trimer re- E:T ratios and receptor densities with the help of a QSP model.
lationship while using the same underlying mechanism, including In summary, well- designed pharmacological experiments in
bell-shaped dose–response relationships. Several publications have combination with fit- for-
purpose mathematical modeling ap-
demonstrated the utility of mechanistic models as exemplified by proaches may be used for the selection of a starting dose for TCEs.
Jiang et al., where the authors captured the in vitro cytotoxicity While such an approach offers a structure to translate from in vitro
data of blinatumomab and extrapolated the effect of target expres- assays to in vivo experiments and the clinic, one must carefully eval-
sion, binding affinity, and E:T ratios to retrospectively calculate uate the inbuilt assumptions when projecting activity for untested
exposure–response relationships in patients (Table 4). Similarly, a scenarios, and continuous validation of the modeling approach
QSP model was able to capture the kinetics of tumor cells and T using multiple data sets is necessary. It is worth noting that the reg-
cells at various E:T ratios and estimate the trimer concentrations ulatory agencies are increasingly acknowledging the value of QSP
required for the minimum biological effect at the physiologically modeling, not only for FIH starting dose selection but also for
relevant conditions for p-cadherin × CD314 (Tables 2 and 4). The clinical decisions including recommended phase II dose (RP2D)
MABEL-based starting doses obtained using traditional exposure- (discussed in ‘Dose escalation, CRS mitigation, and recommended
based calculation and QSP model-estimated trimer-based calcula- phase 2 dose’).
tion were similar. However, in retrospect, the QSP model would
have enabled a much higher starting dose if the dose was selected Projection of human pharmacokinetics
based upon the trimer formation following the first dose rather In addition to the determination of MABEL concentration, es-
than at steady state. Most recently, a QSP model was used to sup- timation of human PK is critical to derive the FIH starting dose.
port FIH starting dose selection of CDR404, a highly potent and For monoclonal antibodies in general and especially for TCEs,
highly specific antibody-based T-cell engager that binds bivalently human PK parameters are empirically estimated from preclinical

CLINICAL PHARMACOLOGY & THERAPEUTICS | VOLUME 0 NUMBER 0 | Month 2024 7

PK parameters by allometric scaling. The disposition of TCEs is to implement novel dose optimization strategies beyond the tra-
influenced by molecular characteristics (e.g., size, FcRn binding, ditional MTD- based dose selection paradigm. The guidance
charge) and is generally well-estimated in NHPs, making them a highlights the importance of exposure–response assessments, dose
relevant preclinical model for human PK estimation. Various stud- adjustments, low-grade adverse events, and relevant specific pop-
ies have demonstrated that fixed exponent scaling on NHP clear- ulations, and encourages the comparison of multiple dosages for
ance can be used for mAbs to estimate human clearance within a efficacy and safety before or during registration trials. The goal of
2× estimation error.18–20 Besides NHPs, the homozygous hFcRn the guidance is to select an optimal dose that not only has maximal
Tg32 mice have been also shown to reliably estimate human PK efficacy but also ensures a favorable safety profile, limiting dose re-
of mAbs21,22 as well as TCEs.10 Recently, Saber et al., 23 reviewed ductions and stoppages during the treatment period.
INDs of 14 TCEs to examine if simple PK modeling (SPM) was The messaging behind Project Optimus is very relevant for
acceptable for FIH dose estimation. The FIH dose based on SPM TCEs, where the selection of doses based up an MTD-based ap-
is determined as a product of estimated human clearance (CLh), proach are not relevant. The dynamic MoA of TCEs and the com-
target plasma concentration ([C]), and dosing interval (tau) and plexity of disease pose several challenges to select an optimal dose
was deemed to be acceptable when the scaling exponent of clear- early in the clinic, requiring more rational dose optimization strat-
ance was fixed to be 0.75. A limitation of simple PK modeling is egies. In addition, physiological features like heterogeneity in dis-
its inability to capture nonlinear pharmacokinetics. TCEs could ease and genetics may result in high inter-individual variability in
potentially undergo TMDD, especially at low doses, depending efficacy and safety. Often, lack of relevant biomarkers and limited
on their interaction with either the target antigen or CD3 recep- follow-up duration also limit assessment of durability of response
tor. Despite being the most relevant preclinical model for human during early trials. Furthermore, TCEs can present a broad thera-
PK estimation, NHPs can fail to capture the complete impact of peutic window (depending on the target) and bell-curve phenom-
TMDD. This may be due to differences between NHPs and pa- enon which makes the MTD-based paradigm not applicable for
tients in key physiological elements like the target burden, target dose optimization. Therefore, for TCEs, it is recommended that
distribution/localization or molecular components like cross- the dose selection is guided by PK, PD, efficacy, and safety data
reactivity, binding affinity, etc. In cases where a TMDD profile collected in early development, across multiple cohorts during dose
is observed in NHPs, deeper mechanistic insights such as target escalation along with the application of MIDD using various tools
turnover, and internalization rates are required to apply a mech- described in later sections.
anistic TMDD model. However, in the absence of such insights,
a Michaelis–Menten model could be employed to capture non- Route of administration (RoA)
linearity and estimate nonlinear PK kinetics in humans. From a Of the 8 approved TCEs, 1 is administered by continuous intra-
safety perspective at the starting dose, capturing nonlinearity may venous infusion (cIV), three following intravenous (IV) infusion,
not be crucial as linear PK profiles tend to overestimate the expo- and four are administered following subcutaneous (SC) injection
sure and are therefore more conservative. However, emerging early (Table 3). Blinatumomab is a tandem scFv-based CD19 × CD3
clinical PK data should be modeled to refine the human PK simu- bispecific T-cell engager (BiTE) which lacks an Fc domain and
lations at higher doses to capture nonlinearity (if any) and provide is currently administered by continuous infusion due to a short
further insights on dose escalations. elimination half- life. The subsequent generation of approved
TCEs are IgG-based and/or have IgG like PK and could be ad-
CLINICAL PHARMACOLOGY CONSIDERATIONS ministered following IV or SC administration. Dosing of TCEs
Project optimus moved toward SC administration, as this RoA results in slower
Clinical pharmacology and early development strategies for on- absorption of the TCE, with reduced maximum concentration
cology therapeutics have recently been influenced by Project compared with IV dosing at equivalent dose levels, which can
Optimus, an initiative by the FDA intended to reform the dose blunt cytokine release and reduce CRS. Of the four TCEs that
optimization and dose selection paradigm in oncology drug de- were approved for SC dosing, epcoritamab is the only TCE to
velopment. While the concept of dose optimization is not new, dose via SC throughout its entire clinical development. The other
recent years have seen increasing engagement between indus- three TCEs initiated trials by dosing IV only to later switch to
try, academia, cancer research centers, and health authorities to SC during the dose-escalation phase before establishing RP2D.
bring renewed focus to pre-market dose optimization strategies The SC RoA is generally considered more patient-compliant and
in oncology drug development. Recent publications including the convenient for long-term use, and evidently, 3 out of the 4 TCEs
white paper from Friends of Cancer Research, 24 presentations at that are approved with IV RoA are currently evaluating SC RoA
ASCO, 25 a review article by the FDA, 26 and ultimately the FDA as part of late development efforts.
draft guidance published in 202327 have detailed the background Given that SC dosing is a preferred RoA for biologics in terms
and importance of Project Optimus. of patient compliance, a common question during development
The essence of Project Optimus lies in selecting an optimal dose is whether FIH studies for TCEs should be conducted directly
during early clinical development, enabling evaluation of safety with SC or start with IV and then switch to SC. There is no one
and efficacy at the optimal dose(s) during registration trials, in- strategy that may fit across targets and indications, and one must
cluding phase II studies that may support accelerated approval for a consider several factors before deciding between SC vs. IV for
novel product. To derive an optimal dose, sponsors are encouraged FIH studies. Mosunetuzumab reported a reduction in grade II

8 VOLUME 0 NUMBER 0 | Month 2024 | www.cpt-journal.com

Table 3 Clinical pharmacology summary of approved CD3-redirecting bispecific antibodies for hematological indications
Clinical data,
Studies to PK in special translational and/or
Generic name support initial Approved dosing Clinical population and clinical PK/PD analysis
(target) Indication RoA approval regimen pharmacokinetics covariates Immunogenicity to support RP2D

Blinatumomab R/R ALL in CIV MT103-104 SU dose of 9 μg Linear over a 2-fold difference Less than 2% of Manageable safety
(CD19)42,51 adults and and followed by TD of 28 μg. dose range from in mean clearance patients treated were profile with a high
children MT103-211 Cycle = 42 days including 5 to 90 mcg/m2/ values between tested positive for MRD-response rate
12 days of treatment- day (approximately patients with moderate ADA, and of patients was observed at a
free period. Treatment equivalent to 9 to renal impairment who developed ADA, dose of 15 μg/m2/
period = 2 cycles of 162 mcg/day) in and normal renal 78% had in vitro day (approximately
induction, 3 cycles of adult patients. The function. However, neutralizing activity corresponds to 28 μg/
consolidation, and up estimated mean the associated day). The Cavg at ss
to 4 cycles of continued systemic t1/2 is variability was very at 28 μg/day ranged
therapy 2.10 hours high within the range. from 500–700 pg/
Effects of severe mL, which exceed the
renal impairment in vitro IC90 value of
are unknown. 470 pg/mL for B-cell
Body surface area suppression in NALM-6
influences the PK, CD19 human B-cell
however, the clinical precursor leukemia
relevance of this effect cell line incubated with
is unknown healthy donor
Tebentufusp HLA-A*02:01- IV IMCgp100-102 SU doses at 20 μg and Linear over a dose Along with ADA, GFR Approximately, A dose of 73 μg was not
(gp100)52 positive adult (15–20 min) and 30 μg on Days 1 and 8, range from 20 to rate was identified 29–33% of treated tolerated. 68 μg was the

CLINICAL PHARMACOLOGY & THERAPEUTICS | VOLUME 0 NUMBER 0 | Month 2024

patients with IMCgp100-202 followed by TD of 68 μg 68 mg. The estimated as the covariate patients tested next tested dose with
unresectable on Day 15. Once every mean systemic t1/2 is with decrease in positive for ADA. manageable safety and
or metastatic week thereafter at 68 μg 7.5 hours clearance in moderate Neutralizing ability improvement in overall
uveal renal impairment unknown. Clearance survival. No clinically
melanoma patients. However, the increased in patients meaningful difference
associated variability with high titer ADAs. was observed between
was very high and no Exploratory analyses weekly vs. daily dosing.
clinical relevance to suggest ADA has no Flat dosing was
efficacy or safety were significant effects on preferred because of
identified clinical activity higher incidences of
AEs in patients with
higher BW
Teclistamab Adult patients SC MajesTEC-1 SU doses at 0.06 and Linear over a dose No key covariates. 1/186 (0.5%) of A positive E-R
(BCMA)41,53 with R/R 0.3 mg/kg on Days range of 0.08–3 The effects of severe treated (SC) patients relationship was
multiple 1 and 4, followed by mg/kg. Teclistamab renal impairment or tested positive for observed, and the
myeloma TD at 1.5 mg/kg on clearance decreases moderate-to-severe ADA. Because of the response rates
Day 7. Subsequent TD over time, with a mean hepatic impairment on low occurrence of approached a plateau
at 1.5 mg/kg weekly maximal reduction pharmacokinetics are ADAs, the effect of close to exposures
or 3 mg/kg bi-weekly from baseline to the unknown ADA on clinical PK and achieved at the RP2D.
thereafter 13th treatment dose of efficacy is unknown The Cavg at ss in both
40.8%. The estimated serum and BM at the
mean systemic t1/2 is RP2D exceeded the
7.5 hours max EC90 observed
from an ex vivo
cytotoxicity in MM
patient samples

(Continued)

9
WHITE PAPER

Clinical data,
Studies to PK in special translational and/or
Generic name support initial Approved dosing Clinical population and clinical PK/PD analysis
(target) Indication RoA approval regimen pharmacokinetics covariates Immunogenicity to support RP2D

Mosunetuzumab Adult patients IV (2–4 h) GO29781 SU doses at 1 and Linear over a dose No key covariates. No ADA observed A positive E–R
(CD20)43,54 with R/R FL 2 mg on Days 1 and 8, range from 0.2 mg to The effects of severe treated patients relationship was
followed by loading dose 60 mg. The estimated renal impairment or (N = 418) ADA. Based observed, and the
of 60 mg on cycle 1 day systemic t1/2 is moderate-to-severe on these data, the response rates
15 and cycle 2 day 1. 16.1 days hepatic impairment on clinical relevance of approached a plateau
Cycle 3 onwards 30 mg pharmacokinetics are ADA on clinical PK and close to exposure
on day 1. Cycle = 21 days unknown activity is unknown achieved at the
RP2D. Two loading
doses at 60 mg
was used to ensure
maximal response
rates in majority of
patients including high
tumor burden and
baseline rituximab
concentrations
Epcoritmab CD19 SC GCT3013-01 SU doses at 0.16 and Exposure increased No key covariates. 2.6% of treated A positive E–R
(CD20)55,56 0.8 mg on Days 1 and 8, more than dose The effects of severe patients (4 of 156) relationship was
followed by full dose of proportionally over the renal impairment or were positive for observed, and the
48 mg on days 15 and dosage range from moderate-to-severe ADA. Because of low response rates
22. QW cycles 2 and 3; 1.5 to 60 mg. The hepatic impairment on occurrence of ADAs, approached a plateau
Q2W cycles 4 to 9: Q4W estimated systemic pharmacokinetics are the effect of ADA close to exposures
cycle 10 onwards t1/2 at 48 mg dose is unknown on clinical PK and achieved at the
22 days efficacy is unknown RP2D. Majority of
initial responses were
observed during the
first 3 cycles (with QW
regimen). Q4W regimen
was successful in
maintaining the
response. QSP
modeling further
informed that the
trimer formation
saturated near RP2D

(Continued)

VOLUME 0 NUMBER 0 | Month 2024 | www.cpt-journal.com

15326535, 0, Downloaded from https://ascpt.onlinelibrary.wiley.com/doi/10.1002/cpt.3439 by University Of Science, Wiley Online Library on [25/11/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Table 3 (Continued)
Clinical data,
Studies to PK in special translational and/or
Generic name support initial Approved dosing Clinical population and clinical PK/PD analysis
(target) Indication RoA approval regimen pharmacokinetics covariates Immunogenicity to support RP2D

Glofitamab Adult patients IV (2–4 h) NP30179 Pretreat with a single Linear over a dose No key covariates. 1.1% (5/448) of A positive E–R
(CD20) [44, 57] with R/R DLBCL 1,000 mg dose of range from 0.005 to The effects of severe treated patients relationship was
or LBCL arising obinutuzumab (Cycle 30 mg. The estimated renal impairment or tested positive for observed for efficacy.
from FL 1 Day 1) 7 days systemic t1/2 is moderate to-severe ADA. Because of the Favorable safety and
before the initiation 7.6 days hepatic impairment on low occurrence of efficacy was observed
of glofitamab (Cycle pharmacokinetics are ADAs, the effect of at the RP2D regimen
1 Day 8). SU doses 2.5 unknown these ADAs on clinical of 2.5/10/30 mg.
and 10 mg on cycle PK and activity is The highest fixed
1 days 8 and 15. Cycle unknown dose studied during
2 to cycle 12–30 mg. the dose-escalation
Cycle = 21 days studies was 25 mg.
The 30 mg target dose
was selected based
on the E-R modeling to
maximize the potential
for efficacy

CLINICAL PHARMACOLOGY & THERAPEUTICS | VOLUME 0 NUMBER 0 | Month 2024

Talquetamab Adult patients SC MonumenTAL-1 SU doses at 0.01 and Linear over a dose The volume of 45/177 (25%) and A positive E–R
(CD20) [58, 59] with R/R 0.06 mg/kg on Days 1 range from 0.005 distribution and 24/130 (18%) of relationship was
multiple and 4, followed by TD to 0.8 mg/kg. The clearance of treated (SC) patients observed, and the
myeloma at 0.4 mg/kg on Day estimated t1/2 is Talquetamab at 0.4 mg/kg weekly response rates
7. Subsequent TD at 8.4 days increased with and 0.8 mg/kg bi- plateaued at or above
0.4 mg/kg weekly or increasing bodyweight weekly developed RP2Ds. Although lower
0.8 mg/kg bi-weekly (40–143 kg). No ADAs. There was no dose of 0.135 mg/kg
key covariates. The identified clinically weekly SC also showed
effects of severe significant effect of activity in responses
renal impairment or ADAs on clinical PK in several participants,
moderate-to-severe and activity fewer participants
hepatic impairment on achieved VGPR or
pharmacokinetics are better, and DOR
unknown tended to be shorter at
these doses compared
with the putative
RP2Ds. The Cavg at ss
in both serum and BM
at the RP2D exceeded
the max EC90
observed from an ex
vivo cytotoxicity in MM
patient samples

(Continued)

11
WHITE PAPER

15326535, 0, Downloaded from https://ascpt.onlinelibrary.wiley.com/doi/10.1002/cpt.3439 by University Of Science, Wiley Online Library on [25/11/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
15326535, 0, Downloaded from https://ascpt.onlinelibrary.wiley.com/doi/10.1002/cpt.3439 by University Of Science, Wiley Online Library on [25/11/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
WHITE PAPER

and grade III CRS events following SC when compared with IV.28

following switch to Q2W.

clinical PK/PD analysis

switch from QW to Q2W

Additional QSP analysis
plateau at the selected

ADA, anti-drug antibody; BW, body weight; cIV, continuous IV infusion; DLBCL, diffused large B-cell lymphoma; E-R, exposure – response; FL, follicular lymphoma; IV, intravenous; MRD, minimal residual disease;
observed between free
elranatamab exposure
Switching from IV to SC results in additional dose-finding cohorts

patients were able to

maintained following
translational and/or

indicated that trimer

maintain responses
and response rates

PK, pharmacokinetic/s; Q2W, bi-weekly; Q3W, once every 3 weeks; Q4W, once every 4 weeks; QW, once weekly; R/R, relapsed/refractory; RoA, route of administration; RP2D, recommended phase II dose; SC,
which appeared to

to tumor ratio was

RP2D. Majority of
to support RP2D

relationship was
to bridge IV dosing schemes to exposure-matched step-up (SU)

in responders
A positive E-R
Clinical data,

and treatment doses via SC route. On one hand, starting an FIH

study following the SC route may save time and resources during
late development by avoiding such additional bridging cohorts. On
the other hand, significant early investments in formulation devel-
opment will be necessary to support SC dosing during FIH trials,
particularly to administer higher doses during the dose-escalation
pharmacokinetics are neutralizing antibodies
ADAs, and 60% (9/15)

clinical PK and activity

against elranatamab.

occurrence of ADAs,
the effect of ADA on
tested positive for

tested positive for

8.9% (15/168) in

phase. Additionally, for targets expressed in the skin and/or drug

treated patients
Immunogenicity

Because of low

is unknown
products that are prone to cause injection site reactions, evaluating
IV instead of SC during FIH studies may mitigate injection site
reactions.
PK profiles following SC administration often lead to a lower
Cmax-to-Ctrough ratio when compared with IV regimens. This may
enable consistent stimulation of T cells and easier maintenance
hepatic impairment on
The effects of severe
renal impairment or
moderate-to-severe
No key covariates.

of static effective concentrations in patients. On the other hand,

population and
PK in special

covariates

relatively higher Cmax-to-Ctrough ratio following IV administration

unknown

may allow sufficient time for T cells to disengage from continuous

subcutaneous; SU, step-up; t1/2, terminal half-life; TD, target dose or treatment dose.References – USPI, FDA and/or EMA BLA assessments.

activation and thereby limit T-cell exhaustion and may also avoid
oversaturation of T cells that can lead to bell-shaped response or
restricted efficacy. However, these concepts remain hypothetical
due to the lack of evidence that demonstrates the impact of con-
range from 6 to 76 mg.
The estimated t1/2 is
Linear over a dose
pharmacokinetics

tinuous T-cell activation or exhaustion on the efficacy and safety

of TCEs in the clinic. In terms of immunogenicity, literature re-
22 days.
Clinical

ports remain inconclusive to draw a correlation between the rate

of ADA incidents and RoA. As an example, a relatively higher in-
cidence of ADA was observed following SC when compared with
IV for JNJ-081 (PSMAxCD3); however, the impact of observed
higher incidence of immunogenicity following SC was considered
32 mg on Days 1 and 4,
followed by TD at 76 mg

Thereafter, bi-weekly at
on Day 7. Subsequent
SU doses at 12 and

inconclusive due to the differences in PK between IV and SC and

weekly TD at 76 mg
through 24 weeks.
Approved dosing

doses tested.29
regimen

76 mg

When switching between IV and SC routes, one must account

for differences in PK to propose a dose that is safe and effective. In
general, since the exposure at a given dose is lower for SC than IV,
the switch from IV to SC can be done at an IV dose that is deter-
mined optimal from a safety and efficacy perspective. Alternatively,
MagnetisMM-3

one can also use population PK modeling to select an SC dose that

support initial
Studies to

approval

leads to similar exposure as the IV dose. However, such analysis

should carefully consider assumptions on the bioavailability fol-
lowing SC and exposure–response relationships for efficacy and
safety following IV. To switch from SC to IV, dose selection may
be informed by PK modeling from SC to IV, and often the starting
RoA

dose for IV is 1–2 cohorts lower than the safe SC dose. Overall,
SC

regardless of IV and SC, the maximum tolerated exposures (respec-

tive Cmax and/or AUC) following treatment doses are generally
expected to be similar. While Cmax is often associated with CRS
Adult patients
Indication

myeloma
with R/R
multiple

risk after a single dose, AUC may be a more relevant metric for cor-
relating with safety outcomes following repeated administration.
Table 3 (Continued)

Additionally, it may be possible for a TCE to be developed for both

IV and SC RoA to give more treatment options to patients and
physicians.
(BCMA) [30, 60]
Generic name

Elranatamab

BW vs. flat dose

(target)

Like mAbs, TCEs are primarily eliminated from the body via
proteolytic catabolism and target-binding mediated intracellular

12 VOLUME 0 NUMBER 0 | Month 2024 | www.cpt-journal.com

degradation. For target-mediated drug elimination (TMDD), MIDD approaches to guide dose escalation (9). Furthermore, one
target and disease burden play a more significant role than body needs to account for target and molecule-specific factors, such as
size. Hence, both flat as well as body size-based (either body on-tumor and off-tumor expression, soluble target engagement,
weight or body surface area-normalized) dosing approaches have and safety margins (based on preclinical studies) while designing
been used in adults for TCEs. Among the approved TCEs, the dose-escalation cohorts.
dosing regimen of most TCEs includes flat dosing, except for te- Determination of maximal dose: The selection criteria for the
clistamab and talquetamab which are administered based on body maximal dose are similar to non-TCE mAbs. The maximal dose in
weight (Table 3). While body size (body weight or body surface a dose-escalation study could be determined based on if a poten-
area) has been identified as a significant covariate for PK of some tial maximal efficacy has been observed, if the maximal tolerated
TCEs (such as blinatumomab, teclistamab, and talquetamab), dose has been identified or a combination of multiple factors (e.g.,
its overall impact on clinical outcomes (exposure vs. response potential benefit/risk ratio, a flat dose–response relationship for
relationship, safety and/or efficacy) has been limited or in some efficacy but dose-dependent-safety, et al.). These approaches are
cases unknown. Additionally, consistent responses with limited further detailed in ‘Recommended phase 2 dose (RP2D)’.
inter-individual variability have been shown after flat dosing with
epcoritamab, mosunetuzumab, glofitamab, and odronextamab, CRS mitigation and step-up dose optimization. CRS is a potentially
thereby suggesting that body size-based dosing might be unwar- serious side effect and a common dose-limiting toxicity associated
ranted.30 Besides, the fixed-dosing approach can have additional with CD3-bispecific T-cell-mediated immune activation. T-cell
advantages including simplified compounding, reduced medica- activation typically triggers an immune response involving the
tion errors, and reduced costs.31 Given the caveats with choosing systemic release of pro-inflammatory cytokines such as IL-6,
BW vs. flat dosing, the decision should be based on observed data. TNFα, and IFNγ. Depending on the magnitude and rapidity
As a general approach, flat dose is an accepted strategy for FIH tri- of release, this cytokine amplification can trigger adverse effects
als for TCEs. For late phase or pivotal studies, emerging early clin- ranging from fever within the first 24 hours, hypoxia, and
ical data should be leveraged and a combination of population PK, hypotension to multiorgan failure and neurological events.32,33
covariate analysis, and exposure–response (for safety and efficacy) Factors related to patient, disease, and prior therapy influence the
analysis should help guide the choice between BW and flat dosing. risk of developing CRS and impact its severity. Tumor burden,
Essentially, in cases where BW is clearly identified as a significant affinity for tumor antigen and its expression level, affinity for CD3,
covariate for both drug exposure and clinical response from early bispecific dose, circulating disease, and pre-existent inflammation
phase studies, body weight-based dosing could be used, otherwise, can influence a CRS outcome.32,34 Thus, minimizing the incidence
flat dosing can be implemented for pivotal studies. and mitigating the severity of CRS is a critical consideration for
any CD3-bispecific clinical development program.
Dose escalation, CRS mitigation, and recommended With CRS, the priming effect is a phenomenon common to
phase II dose CD3-bispecific agents, with the most severe symptoms observed
following the first dose and a subdued cytokine response observed
Dose escalation. Dose-
escalation strategy for TCEs generally with subsequent doses of similar or higher magnitude. The kinetics
involves three main considerations: escalation of step-up doses, of this phenomenon are exploited to mitigate CRS incidence. The
escalation to treatment doses, and determination of maximal dose. use of step-up dosing is one approach, which involves administer-
Escalation of step-up dose: As CRS is a commonly observed ing one to two smaller initial doses a few days apart followed by the
dose-limiting toxicity for TCEs, step-up dosing strategy is usually intended treatment dose (Table 3). This approach has been used
implemented for CRS management (detailed further in ‘CRS mit- to optimize the dosing regimen for all currently approved CD3-
igation and step-up dose optimization’). Given that the starting bispecific agents. The selection of the optimal step-up dose(s) and
dose(s) selected based on the MABEL approach is generally below regimen has typically been performed empirically and has been a
the finally selected step-up doses, single patient cohorts could be challenge given many possible combinations of dose and dose fre-
considered at early cohorts until ≥ grade II CRS events were ob- quency. For instance, one common approach is to select a dose level
served. For early cohorts, threefold dose increments followed by which minimizes the CRS grade ≥2. The utilization of one or two
slower increments (twofold or lower) could be considered based on step-up doses is primarily determined by how well-controlled the
target and molecular properties and clinical safety. incidence and severity of CRS is with each approach. A recent ex-
Escalation to treatment dose: Escalation to treatment or effec- ample of clinical evaluation of one-step vs. two-step priming dose
tive doses is generally done in multi-patient cohorts (e.g., 3–6) is for elranatamab, where two step-up priming doses (Day 1 and 4)
guided by safety signals using Bayesian optimal interval (BOIN) resulted in a lower incidence of any grade and grade ≥2 CRS vs. a
designs. In addition, exposure profiles (e.g., from nonlinear to lin- single priming dose (Day 1).35 Testing different regimens guided
ear PK, exposures at specified dosing frequencies), efficacy signals by observed preclinical data and clinical data from initial phase I
(e.g., objective response rates by cohort), clinical biomarker infor- dose-escalation cohorts can be useful in the selection of a step-up
mation (e.g., saturation in RO, T-cell activation and proliferation) regimen.36 Recent examples of the use of model-g uided selection
(16), and relevant preclinical information (e.g., observed effica- of step-up doses using PK/PD and QSP models offer an attractive
cious concentrations in in vitro or in vivo pharmacology experi- opportunity to expedite the selection of an optimal step-up dosing
ments) should be considered and assessed jointly with the help of regimen and should be utilized, where feasible.37,38

CLINICAL PHARMACOLOGY & THERAPEUTICS | VOLUME 0 NUMBER 0 | Month 2024 13

The use of a pre-medication cocktail of a corticosteroid (dexa- PK simulations could be employed to capture different scenarios
methasone), an antihistamine (diphenhydramine) and an an- like alternate dosing regimens and restarting therapy after a
tipyretic (acetaminophen or paracetamol) is a recommended dosing interruption. For example, PK simulation has been used in
approach to blunt the impact of cytokine release, and typically the selection and justification of a “repriming window” for TCEs.
employed for the initial step-up dose(s) and first treatment dose.39 As described previously, a step-up-dosing strategy is commonly
For the FIH study, it is recommended that patients are hospitalized applied for TCEs to mitigate the risk of CRS. Dosing delays or
for a duration of time (e.g., up to 48 hours) for the initial step-up interruptions due to AE or other reasons, particularly during the
and first treatment doses, to monitor the CRS risk and allow early step-up periods or early on during treatment can compromise the
clinical intervention. These durations can be updated based on effectiveness of step-up-dosing and put patients at elevated risk
developing clinical data and inform label recommendations.39 of CRS. An extended dosing delay may lead to an unacceptable
Considerations for the RoA and switching routes are discussed in increase in CRS risk and restarting step-up-dosing may become
‘Route of administration (RoA)’. necessary. PK simulation can be used to determine and justify an
Clinical management of CRS follows a graded approach ranging acceptable “repriming window” and such an approach has been
from temporary to permanent discontinuation of treatment, IV generally accepted by the FDA for TCEs to inform the drug
fluids, and oxygen supplementation for lower grades to monitor- label. Generally, the “repriming window” is selected based on
ing of organ function in the ICU for higher grades or prolonged PK simulation of the dosing delay and time it takes for the drug
CRS.34 Tocilizumab is an anti-IL-6 antibody that is commonly concentration to fall below a certain specified concentration
recommended for use with higher grade CRS and preferred over threshold (typically exposure during or after step-up-dosing is
the use of corticosteroids, unless the CRS is refractory to anti-IL6 chosen as the threshold). In addition to the PK-based approach,
therapy or severe neurotoxicity is observed, since tocilizumab does a longitudinal PK/PD modeling-based approach can be used to
not cross the blood–brain barrier. In such instances, corticosteroids address the dosing schedule and repriming window selection,
may be considered as the next line of intervention for CRS.33,40 which will be discussed in later sections.

Recommended phase II dose (RP2D). Dose expansion and RP2D Exposure–response analysis. Exposure–response (E–R) analysis
for TCEs are guided by several MIDD approaches. This section is commonly conducted for TCEs to guide on two critical
provides an overview of MIDD approaches and our perspective outcomes. First, to derive E–R relationships in phase Ia studies,
toward RP2D determination with the help of case studies. In from the starting dose to treatment doses, assessing the impact
general, MIDD is an iterative process, it should start in preclinical of increasing exposures on efficacy or safety. Second, to derive
stage and is continuously refined as more data becomes available. E–R relationships within the patient population at the RP2D
Different modeling approaches might be applied in different to characterize the variability observed in the clinical response.
stages of clinical development. Clinical efficacy signs for TCEs are often observed early
during the first one or two disease evaluation periods (e.g., within
Model informed drug development approaches and case studies. 1–3 months of treatment). Therefore, PK metrics such as the area
Population pharmacokinetics. In principle, the population under the curve (AUC) or trough concentrations (Ctrough) used for
PK modeling approach for TCEs is similar to monospecific E–R analysis should be relevant to the early repeat doses. Typical
antibodies. The PK of TCEs is often characterized using a one or clinical efficacy end points used for E–R analysis to guide dose ex-
two-compartment PK model with time-dependent and/or time- pansion and RP2D include overall response rate (ORR), partial
independent clearance. Drug exposures during the escalation response (PR), and complete response (CR). During the early part
phase can span a broad range (e.g., up to 5 log units). The broad of the FIH trial, typically there is limited data on other end points,
dose range covers a low MABEL-based starting dose to the CRS such as progression-free survival, duration of response, and over-
mitigating step-up doses, all the way to a typically higher treatment all survival, thus precluding inclusion of these end points for dose
dose where maximal efficacy and/or maximal tolerability (MTD) expansion and RP2D decisions. E–R analysis may also be used to
is observed (14, 15). Elimination could be dominated by a characterize safety observations to derive an exposure threshold for
saturable TMDD mechanism at lower doses and by non-specific tolerability. Such analysis can be helpful in determining the opti-
elimination at higher treatment doses. In addition, several TCEs mal dose/exposure that leads to maximal efficacy with minimal
have also exhibited a time- dependent decrease in clearance safety events.
following repeat doses. The mechanism for time- dependent For example, E–R analysis played a pivotal role in determin-
clearance is unknown and could be a result of multiple factors, ing the RP2D/regimen for mosunetuzumab. E–R analysis indi-
such as on-and off-tumor target expression, tumor burden, cated that clinical responses (ORR and CR rates) increased with
and cachexia. Population PK modeling and simulation play an mosunetuzumab exposure (AUC0-42) and approached a plateau.41
integral role in the clinical development of TCEs, providing EC90 was estimated using a logistic regression Emax model, and a
exposure estimates for exposure–response analysis, PK/PD and shift to the right (increasing EC90) was observed in patients with
QSP modeling. Importantly, population PK analysis provides high baseline tumor size. A similar shift was also seen in patients
information on inter- individual variability in PK through with high residual concentrations of CD20 targeting antibodies
covariate analysis that informs follow-up analyses to characterize (e.g., rituximab) at baseline. Furthermore, the E–R analysis was
the efficacy and safety across patient population. Additionally, used to demonstrate that for 95% of patients receiving the RP2D

14 VOLUME 0 NUMBER 0 | Month 2024 | www.cpt-journal.com

regimen, the average exposure (AUC0–42) would be at or near the – CRS E–R analysis demonstrated a diminishing relationship be-
plateau (above EC90), with the exception being the remaining 5% tween IL-6 peak concentration and CRS incidence over time (over
of the patients with high levels of circulating CD20 targeting an- the four dosing periods in cycle 1) indicating that the step-up dos-
tibodies and at the highest (95th) percentile of baseline tumor ing regimen attenuated the incidence of CRS.42 However, applica-
size. E–R analysis was also extended to characterize safety events, tion of E–R analyses in this manner has limitations. By performing
including CRS, neurological, and other adverse events (≥grade E–R analyses separately for each dosing period, the cumulative ef-
3). E–R analysis for CRS indicated a positive correlation in fixed- fect of the step-up dosing regimen and interactions between the
dosing cohorts and reduced effects of exposure following the im- step-up doses and the target dose cannot be accounted for. As such,
plementation of step-up dosing. However, the final step-up dosing a longitudinal PK/PD modeling-based approach may be superior
regimen was chosen empirically based on the review of clinical for step-up-dosing regimen selection and optimization, which will
data across dose escalation-cohorts.41 While there appeared to be be discussed in the next section.
a slight positive trend for exposure and neurological events, the ob- Overall, E–R analyses are highly effective to assess the relation-
servations were statistically insignificant for neurological adverse ships between exposure and efficacy or safety observations to bet-
events. Additionally, higher exposure did not appear to lead to an ter understand the observed variability in response observed in the
earlier onset of other grade ≥3 adverse events. clinic. However, exposure alone may not be sufficient to address
Selection of the exposure metric used in exposure–response anal- the differences between responders and non-responders. This is
ysis can significantly influence the result and interpretation of the due to a multitude of factors like T-cell dynamics, tumor hetero-
relationship between drug exposure and response. There is no one geneity, and growth rates that play a key role in impacting response
universally applicable exposure metric that works for all situations. and relapse rates. Such assessments are better facilitated with the
For example, mosunetuzumab used day 0–42 AUC as the expo- help of QSP models, as detailed in the following section.
sure metric while epcoritamab used day 0–28 AUC as the expo-
sure metric. The choice of appropriate exposure should be based on PK/PD and quantitative systems pharmacology modeling. PK/
consideration of the type and time of onset of the response variable PD or QSP models can be used to address the mechanistic
investigated, biology associated with the disease, and potential bias questions pertaining to clinical study design and interpretation
that may be introduced, among other considerations. For example, for TCEs that are beyond the scope of E–R analysis. In general,
when investigating the relationship between drug exposure and PK/PD modeling refers to “top- down” analysis of datasets,
early onset adverse events, such as CRS, exposure metric should with subsequent extrapolation to different experimental
be chosen to represent the AE onset period. When investigating conditions or across species. QSP modeling refers to “bottom-
the relationship between exposure and efficacy in diffused large up” application of modeling, integrating data from disparate
B-cell lymphoma (DLBCL), an aggressive form of lymphoma, an sources, including literature, preclinical experimental, and
exposure metric that covers the early exposure period may be more clinical data to examine the relationships between the drug,
appropriate. In contrast, an exposure metric that covers a wider ex- the biological system, and the disease process.13 Since TCEs
posure period may be more appropriate for follicular lymphoma are complex molecules with many variables, this type of
(FL) exposure-efficacy analysis, given its less aggressive nature. modeling can be useful to address the limitations with clinical
However, selection of an exposure metric that covers a wider period datasets and to bridge the gap between patient populations.
can introduce bias into exposure–response analysis. Dosing delay, PK/PD modeling has served as an impactful tool across several
early patient dropouts, and repriming can introduce bias in the cal- TCEs, providing a rationale for observed exposure-to-efficacy rela-
culated exposure metric and can significantly impact the result and tionships in the clinic and supporting RP2D and regimen determi-
interpretation of the relationship between drug exposure and re- nation. For example, in the case of teclistamab, a PK/PD analysis
sponse. Integrated joint modeling of tumor growth inhibition, sur- indicated that the therapeutic exposure range was between the
vival, and adverse events driven by longitudinal PK may have utility mean EC50 and maximum EC90 of the ex vivo cytotoxicity assay
in the context of TCEs. This approach bypasses the issue of select- (Table 4).43 This assay used frozen PBMCs from patients with
ing a single exposure metric, and replacing it with longitudinal ex- multiple myeloma, incubated with purified T cells from healthy
posure. In addition, this approach allows for integrated assessment donors at a 1:1 E:T ratio. Bone marrow PK was measured, reflect-
of efficacy and safety for the selection and optimization of dose. ing tumor concentrations, and found to be similar to the drug con-
In addition to informing treatment dose for RP2D, E–R analy- centration in serum, enabling direct correlations between serum
sis could be explored to guide step-up dosing regimen selection. In PK and cytotoxic response. Clinical PK modeling and simulation
the case of epcoritamab, E–R analyses were performed to assess the indicated that the exposure achieved in patients following a subcu-
relationship between exposure (Cmax) during the first four doses taneous dose of 0.72 mg/kg reached the maximum observed EC90
during the step-up-dosing period in cycle 1 and the risk of CRS. value from the ex vivo cytotoxicity assay. A weekly subcutaneous
There was no increasing risk of CRS associated with higher expo- dose of 1.5 mg/kg was selected as the RP2D based on a totality
sure during each of the four dosing periods at the selected RP2D of data from efficacy, safety, PK, and PD considerations. Similar
regimen, indicating the suitability of select step-up-dosing at mit- in vitro-based translational PK/PD modeling approaches have
igating the CRS risk (unpublished). In addition, the relationship been found to be applicable for blinatumomab and talquetamab.
between exposure to cytokine and CRS risk during each dosing Alternatively, in the case of odronextamab, translational PK/PD
period can be explored via E–R analyses. For epcoritamab, IL-6 analysis indicated that exposures achieved at the efficacious dose

CLINICAL PHARMACOLOGY & THERAPEUTICS | VOLUME 0 NUMBER 0 | Month 2024 15

Table 4 Review of selected published PK/PD and QSP modeling approaches to guide development of CD3 T-cell engagers
(TCEs)
Target/Product/
Indication Modeling approach/key features Application Key Findings
Semi-mechanistic/mechanistic/QSP modeling
Multiple CD19 Mechanistic model with TBEPC complex To simultaneously characterize Model-based simulations projected
(Blinatumomab) linked to target cell killing in vitro cytotoxicity data across bell-shaped curve for TBEPC under
in ALL, P-cadherin Model includes binding affinities to CD3 different assay conditions (e.g. E:T ranges of different system-and drug-
(LP-DART), CD33 and target, levels of CD3 and target on ratios, time points) and for multiple specific parameters
(AMG-330), FLT3 cell surface, concentrations of effector molecules TBE model-estimated blood and bone
(FLT3xCD3), CD20 and target cells, and other relevant To evaluate impact of different marrow ER curves for blinatumomab
(CD20xCD3)7 physiological parameters system-and drug-specific param- were consistent with clinical findings
eters on ER relationship (complete peripheral B-cell depletion at
To retrospectively estimate effica- priming (9 μg/day) and full dose (28 μg/
cious dose for blinatumomab in the day)) whereas near complete depletion
clinic based on in vitro cytotoxicity in bone marrow only with full dose (and
data not priming dose) even at low baseline
E:T ratios in ALL patients
P- cadherin Mechanistic PK/PD model with trimo- To characterize in vitro kinetic Estimated MABEL human dose based
(LP-DART) lecular synapses driving T-cell prolifera- cytotoxicity data across different on mechanistic PK/PD approach was
Solid tumors13 tion and tumor cell killing assay conditions (e.g., drug con- 1.9 ng/kg/week with a 1-hour infusion
Model accounts for physiological centration, E:T ratios) and estimate (estimated tumor synapse concentra-
parameters including binding affinities relevant EC synapse (EC20, syn) tion at steady state ~ in vitro EC20, syn)
to CD3 and target, levels of CD3 and concentration for killing as in vitro PK-driven approach based on in vitro
target on cell surface, concentrations MABEL cytotoxicity or IL-6 (Cmax < in vitro EC20)
of effector in blood and tumor and Mechanistic in vivo PK/PD model assay yielded similar starting dose
target cells in tumor, soluble P-cadherin was used to estimate human ~1.5 ng/mg/week whereas RO-based
plasma concentration as well as drug MABEL dose based on in vitro approach (Cmax < EC10, RO) resulted in
PK in blood and tumor for in vivo model MABEL synapse concentration and much higher starting dose estimates
human PK estimations (360 and 8,300 ng/kg/week based on
Mechanistic PK/PD approach P-cad and CD3 RO, respectively)
compared with other orthogonal
approaches
P- cadherin Translational QSP model linking trimer Translational model using nonclini- TSC values were estimated to be
(PF-06671008) concentration to tumor cell killing cal data (in silico, in vitro and in 0.0092–0.064 pM based on various
Solid tumors [61] Model includes PK in central compart- vivo mouse efficacy data) to esti- mouse models that translated to ~0.1
ment and tumor, T-cell distribution and mate the efficacy in humans using to 1 μg/kg human dose
proliferation in tumor, T-cell and tumor TSC, that is, minimum efficacious However, the human active clinical
cell binding trimer concentration dose estimates and efficacy estima-
Model-based simulations to evaluate tions were severely impacted by target
impact of different parameters (e.g., levels of P-cadherin and E:T ratio
soluble P-cadherin) on efficacy assumptions
Multiple Next-generation model with two-step To better characterize in vitro cy- The 2D model estimated trimer forma-
CD19 binding – 1st step 3D binding to form totoxicity data across different E:T tion to be less sensitive to the change
(Blinatumomab), dimers; 2nd step with trans-avidity ratios and for multiple molecules in cell densities that better character-
FLT3 (7370), and modeled as a 2D process and To retrospectively estimate active ized in vitro cytotoxicity data across E:T
CD33 (AMG- compared with 3D binding model to dose for blinatumomab in the clinic ratios
330), GPRC5D form trimers based on in vitro cytotoxicity data The 2D model estimated active dose of
(Talquetamab)35 blinatumomab at ~21–24 μg/day which
is comparable to the approved dose
28 μg/day. In contrast, 3D model esti-
mated ~0.006–0.0068 μg/day which is
~1,000x lower than the approved dose
CD20 Translational semi-mechanistic mPBPK To characterize preclinical (monkey) Model-based estimations showed
(epcoritamab) /PD model that characterizes PK/ and clinical data (phase I/II study) trimer formation to plateau at
B-cell biodistribution of epcoritamab, trimer including PK, biomarker, tumor, and 48–96 mg doses and suggested
lymphomas34 formation, and tumor response response optimal responses at 48 mg in DLBLC
Key mechanisms include lymphocyte Preclinical to clinical translation and FL which was comparable to ER
turnover and trafficking, binding kinet- and model-based simulations to es- analysis
ics, T-cell and B-cell dynamics, T-cell timate trimer formation and tumor ER analysis revealed flat relationship
activation, target cell (B-cell and/or response rates in DLBCL and FL between drug exposure and CRS within
tumor) killing patients and guide RP2D selection the dose range evaluated
mPBPK modeling framework previously Model-based estimations supported
developed for mAbs leveraged to de- the selection of 48 mg dose as RP2D
scribe PK and disposition of epcorita- along with totality of evidence and other
mab to various tissue compartments analysis/approaches
(Continued)

16 VOLUME 0 NUMBER 0 | Month 2024 | www.cpt-journal.com

Table 4 (Continued)
Target/Product/
Indication Modeling approach/key features Application Key Findings
CD20 Translational QSP model with T-cell and To characterize preclinical (cyno Step-dosing or dose fractionation was
(Mosunetuzumab) B-cell in vivo dynamics and interactions studies) PK and PD endpoints estimated to reduce systemic cytokine
CD19 with CD3 bispecifics in five different (including T-cell and B-cell dynam- levels or T-cell activation with minimal
(Blinatumomab) physiological compartments (blood, ics, T-cell activation, and cytokine impact on efficacy which informed
DLBCL/NHL 29 bone marrow, lymph node, spleen, levels) Phase I dose-escalation design for
tumor) QSP model using virtual cyno Mosunetuzumab
Key physiological mechanisms incorpo- translated to humans with blinatu- Prospective model-based clinical
rated, such as T-cell subsets, thymic momab data in ALL including PK simulations matched clinical data with
generation, apoptosis, activation and and PD (T-cell activation, B-cell cycle-1 step-up dosing regimen mitigat-
proliferation, de-activation, inter- killing) endpoints ing CRS in the clinic
compartmental trafficking, along with Similarly, virtual cyno and human Maximal induction of T-cell activation
others models estimated cytokine levels and systemic IL-6 elevation occurred
PK of CD3 bispecifics and drug (e.g., IL6) and clinical efficacy with after 1st dose of cycle 1
concentration- dependent T- cell activa- blinatumomab in NHL, subse- Step-up dosing can enable administra-
tion, cytokine production, and target quently used to estimate the same tion of higher target doses without MTD
cell killing for different dosing regimens of
mosunetuzumab
CD20 Previously developed QSP model Systems-based digital twins’ Model estimated lower exposure
(Mosunetuzumab) (Hosseini 2020) that further incorpo- approach used to characterize thresholds for responses in patients
NHL38 rated different dosing regimens and clinical dose/exposure–response with indolent subtype as compared with
patient heterogeneity relationship during dose escalation other aggressive NHL types which was
Digital twins’ approach to form virtual and identify determinants of clinical due to the differences in tumor charac-
population to incorporate relevant clini- responses teristics such as tumor growth rate and
cal variability in biological, pharmaco- baseline T-cell infiltration
logical, and tumor-related parameters Model-b ased simulations also identi-
fied tumor parameters (size, growth
rate, baseline T-c ell infiltration) and
parameters characterizing the drug ef-
fect on T-c ell activation and target cell
killing to govern clinical response
Other/Empirical PK/PD approaches
CD20 PKPD Model (in NHPs) including two- Translational PK/PD model to After human scaling, accounting for
(Glofitamab) compartment PK model with linear characterize PK and PD (cytokine species difference from the in vitro
B-cell and nonlinear elimination, indirect release) relationship in NHPs cytokine assays, and applying a safety
lymphomas9 response PD model with transit com- Translational PK/PD model scaled factor of 10, the starting dose for
partments for delayed cytokine release to humans for starting dose selec- humans was estimated at 5 μg
(IL- 6, IFN- γ, and TNF-α), and B-cell tion (at which the levels of IL-6, PK/PD modeling approach resulted in
dynamics in blood (impacting PK/PD most sensitive cytokine stays 30X higher human starting dose esti-
relationship) below a predefined threshold of mates as compared with conventional
PK parameters were mostly scaled to 600 mg/mL) in vitro-based MABEL approach (using
humans using allometric scaling while EC50 value from tumor cell lysis model
PD parameters were assumed to be in the most sensitive cell line)
similar
Species difference (70-fold factor) in
cytokine release was accounted for
from relevant in vitro assays for human
dose estimation
BCMA Nonlinear mixed-effect modeling Characterize human PK from low- PK simulations prospectively estimated
(teclistamab) including 2-compartmental PK model dose cohort in phase I study QW 0.7 mg/kg IV and 0.72 mg/kg SC to
multiple with linear clearance Developed PK model was prospec- yield serum trough concentrations ~ ex
myeloma33 tively used to simulate PK profiles vivo EC90
for higher doses with trough serum QW 0.2 and 0.72 mg/kg IV and 0.72
concentrations higher than the ex and 1.5 mg/kg SC were active in the
vivo EC90 value to estimate thera- clinical study with 1.5 mg/kg QW SC
peutic range selected as RP2D
Bone marrow PK (tumor concentrations)
were comparable to serum PK

(Continued)

CLINICAL PHARMACOLOGY & THERAPEUTICS | VOLUME 0 NUMBER 0 | Month 2024 17

Table 4 (Continued)
Target/Product/
Indication Modeling approach/key features Application Key Findings
PBPK modeling
CD19 PBPK modeling framework (Simcyp) To evaluate impact of Model estimated suppression of he-
(blinatumomab) including CYP enzymes (CYP3A4, blinatumomab-mediated cytokine patic CYP450 by <30%, impact of IL-6
NHL45 CYP1A2, and CYP2C9 based on elevations (IL-6) on the expression on substrates of CYP3A4, CYP1A2, and
in vitro results) and IL-6 cytokine and activity of CYP 450 enzymes CYP2C9 was within twofold and limited
(based on in vitro and in vivo data of to <1 week
IL-6 -m ediated suppression of CYP Model suggested duration of cytokine
enzymes) elevation as compared with levels as a
IL-6 kinetics (zero-order input rate and major determinant of CYP suppression
first-order elimination) and drug PK pa- Overall transient elevation of cytokines
rameters (total CL, Vd,ss) were included with blinatumomab treatment had low
for simulations DDI potential
In vitro experimental data in human
hepatocytes, demographic and physi-
ological parameters from specific pa-
tient population, and clinical data (PK,
longitudinal IL-6) are used to develop
and calibrate the model
2D, Two-Dimensional; 3D, Three-Dimensional; ALL, Acute Lymphoblastic Leukemia; CD, Cluster of Differentiation; Cmax, Maximal plasma concentration;
CRS, Cytokine Release Syndrome; CYP450, Cytochrome P450; DDI, Drug–Drug Interaction; DLBCL, Diffuse Large B-Cell Lymphoma; E:T – Effector-to-Target
Cell Ratio; EC, Effective Concentration; ER, Exposure–Response; FL, Follicular Lymphoma; FLT3, fms related receptor tyrosine kinase 3; GPRC5D, G-Protein
coupled Receptor family C group 5 member D; IL-6, Interleukin-6; MABEL, Minimal Anticipated Biological Effect Level; mPBPK, minimal Physiologically-Based
Pharmacokinetics; MTD, Maximum Tolerated Dose; NHL, Non-Hodgkin Lymphoma; PK/PD, Pharmacokinetics/Pharmacodynamics; QSP, Quantitative Systems
Pharmacology; RO, Receptor Occupancy; TBEPC, Target cell-Biologics-Effector Cell Complex per Target Cell; TSC, Tumor Static Concentrations.

in the xenograft model in mice were associated with the efficacious preclinical experiments and clinical outcomes by accounting for
exposures observed in patients.12 differences in tumor and T-cell dynamics in patients and it also
Longitudinal PK/PD modeling calibrated to clinical data can be leverages physiologically-based pharmacokinetic (PBPK) model-
leveraged to aid in step-up-dosing optimization, “repriming win- ing principles to describe drug biodistribution and exposure in rel-
dow” identification and justification, and dosing schedule selec- evant tissue such as tumor for efficacy or normal tissue for toxicity.
tion. In the epcoritamab example, a repeated time-to-event (rTTE) For example, in the case of epcoritamab, a QSP model was devel-
model was developed using the observed clinical data by linking oped using in vitro binding and NHP PK/PD data, including PK,
the risk of CRS to longitudinal PK by incorporating PK-driven T-cell, and B-cell dynamics (Table 4).45 The model was further cal-
increase in the hazard of CRS event, onset and tolerance dynamics ibrated using clinical PK, patient biomarker data (T-cell and B-cell
of epcoritamab exposure affecting the hazard of CRS events em- count), and clinical response data to successfully support RP2D se-
pirically.44 Simulations were performed to mimic various step-up lection. The model described the binding of epcoritamab to CD3
dosing regimens to narrow the alternate step-up dosing regimens and CD20 on T cells and B cells, respectively, and projected trimer
to evaluate in the clinic to further optimize the step-up dosing formation within dose-escalation cohorts. The model-estimated
regimen for epcoritamab. Although this approach is more bio- clinical response rates and trimer formation that were used to
logically relevant, the underlying mechanisms of cytokine release identify the biologically effective dose of epcoritamab. Model esti-
and CRS within the E–R relationship require further elucidation. mations indicated plateauing of trimer formation within the dose
Additional validation and reverse translation studies with diverse range of 48–96 mg, with a decrease at higher doses of 192 mg due
TCEs are necessary for broader implementation. to the bell-shaped response that is typically expected for TCEs. In
While in vitro or in vivo-based PK/PD modeling is a powerful parallel to QSP, traditional E–R analysis also concluded that the re-
tool, it is heavily reliant on the reliability of the established preclin- sponse plateaued by the 48 mg dose, and therefore, no further dose
ical models in estimating clinical outcomes. This may depend on escalation was implemented. Thus, the 48 mg treatment dose was
the tumor cell line selected, the immunity of mouse, and several selected as the RP2D for registration trials. A recent publication
experimental conditions; some models may have poor translatabil- showed an improved approach of modeling the second binding
ity. TCEs pose unique translatability challenges attributed to the step of the trimer formation as a two-dimensional process instead
multiple binding domains, lack of comprehensive data on target of three-dimensional process, which allows the model to more ac-
expression in tumors, tumor heterogeneity in patients, and T-cell curately translate from in vitro cytotoxicity data to estimate the
dynamics. Therefore, as also discussed in the FIH sections of this efficacious dose of blinatumomab (Table 4).16 Finally, although
publication, preclinical studies should be carefully designed, and not a CD3 T-cell engager, but a relevant and most recent example
the results should be interpreted with caution, with continuous for T-cell engager used QSP model to estimate the efficacious dose
reverse translation efforts for each indication across molecules. for ATG-101, a 4-1BB × PD-L1 bispecific.46 The model was ini-
QSP models have an added advantage to bridge the gap between tially calibrated to in vitro binding data as well as in vitro functional

18 VOLUME 0 NUMBER 0 | Month 2024 | www.cpt-journal.com

assays. The in vitro model was subsequently translated to human. approaches are now commonly being applied across TCEs to gain
A bell-shaped dose–response was estimated for trimer formation mechanistic insights on the RP2D decisions (17).
as expected for this class of molecules. To maximize efficacy, both As guided by Project Optimus, it is important to select a dose
maximum trimer formation and maximum PD-L1 RO are desired. that is efficacious and yet well tolerated in patients. For TCEs, an
Based on these criteria, an optimal dose of 2 mg/kg was estimated MTD-based approach may not be suitable for many targets, and
by the QSP model, and this information was used to guide the clin- one must also account for the potential negative impact on efficacy
ical study design.46 due to bell-shaped response as described earlier. It is recommended
Trimer formation that determines the efficacy of TCEs is im- that the final RP2D decisions are based on appropriate modeling
pacted by multiple factors including target expression and dou- analysis and observed clinical efficacy and safety data in backfill
bling rates, and TCE and T-cell concentrations at the tumor site. and/or dose expansion cohorts (at least 2 or more as appropriate)
As discussed previously, trimer formation in the preclinical setting in the relevant patient population that will be enrolled for late de-
often exhibits a bell-shaped curve; however, there has been limited velopment studies. In general, early dose-escalation cohorts that
evidence for bell-shaped response in the clinic to date. This is likely is focused on evaluating safety include fewer patients per cohort
a result of the large variability observed in tumor and T-cell dy- (e.g., 2–6 patients) in a heterogenous disease population (e.g., lym-
namics in the patient population that could result in highly variable phoma), and the dose expansion cohorts include specific target
concentration for peak response and broader bell-shaped response population to evaluate safety and efficacy (e.g., FL and DLBCL). A
that is generally not evident at the population level. Despite that preliminary exposure–response analysis is generally done with the
bell-shaped dose–response was likely the reason in a recent study data from dose escalation to select 2–3 regimens for expansion and
with tarlatamab (DLL3xCD3)47 where greater efficacy (objective then a combined analysis using data from both dose escalation and
response rates and progression-free survival) was observed at the expansion to determine the RP2D regimen by accounting for any
10 mg dose when compared with 100 mg. differences due to the underlying disease characteristics. Although
In addition to supporting RP2D selection, QSP models can the FDA guidance on dose optimization recommends the use of
aid in capturing patient heterogeneity to identify biological deter- randomization for dose expansion cohorts to guide RP2D when
minants of clinical response. As in the case of mosunetuzumab, a appropriate, there may not be added scientific benefit to implement
digital twin-based virtual population approach was implemented randomized cohorts in most cases to determine RP2D. Instead,
to represent variability in biological, pharmacological, and tumor- backfill cohorts can be utilized to generate clinical data in relevant
related parameters from the clinical study (Table 4). Simulations dosing groups and the RP2D determination should be backed by
indicated lower exposure thresholds for indolent compared with a strong scientific rationale based on exposure–response analysis,
aggressive non-Hodgkin’s lymphoma subtypes because of lower PKPD modeling, QSP modeling, and clinical data. Therefore, it
inferred rates in tumor proliferation and T-cell infiltration in the may be beneficial to have early regulatory interactions to align on
corresponding digital twins.48 The model also identified underly- the RP2D selection strategies with health authorities during early
ing potential biological differences in clinical responders and nonclinical development. While choosing doses for expansion cohorts,
responders, attributing them to tumor parameters, such as size, one must carefully evaluate PK, efficacy, and safety relationships
proliferation rate, and baseline T-cell infiltration. Additionally, to avoid dosing multiple patients at sub-therapeutic (low end) or
it suggested that the expansion of pre-existing T cells within the toxic doses (high end). The FDA has also released an oncology
tumor, rather than a systemic influx of expanded T cells, drives the dosing tool kit that is designed to serve as a resource to support in
anti-tumor activity of mosunetuzumab. The mechanisms of resis- decision-making regarding dosage optimization.50
tance to TCE therapy were also explored using a QSP model fo-
cused on the formation of immunological synapses. Dosing regimen considerations
Overall, QSP models that account for cancer cell heterogene- Dosing regimen for TCEs depends on a combination of factors
ity are potentially better equipped to estimate dosing regimens ca- including half-life and time-dependent clearance properties of the
pable of reducing resistance and preventing relapse.49 Employing TCE and observed efficacy profile including time to best response
this QSP approach, alongside other methods, has the potential and depth of response. It has been shown across several TCEs that
to support the optimization of dosing regimens and combination the first evidence of response is often observed within the first 2
strategies, thereby enhancing long-term treatment outcomes. One months after TCE administration, and exposures during the first
must understand that such estimations using a QSP model gener- few weeks (42 days for mosunetuzumab and 28 days for epcori-
ally come with several assumptions, calibrations, and uncertainties. tamab) play a critical role to derive exposure–response relation-
Therefore, outputs derived from the QSP model should always be ships.45,51 Therefore, the dosing regimen for most of the approved
interpreted with caution by the subject matter experts. TCEs follows a relatively frequent administration at the start of
the treatment followed by a less frequent administration based on
Summary on RP2D selection the observed efficacy (Table 3).
As described in the previous section, RP2D selection for TCEs are Some of the approved TCEs including blinatumomab,
guided by various MIDD approaches and the totality of clinical mosunetuzumab, and glofitamab are administered over a fixed
safety and efficacy data from all enrolled patients in the FIH study treatment duration ranging from 6 to 12 months.52–54 Response to
at the time of decision. In addition to population PK and E–R treatment has been observed to be durable and extended beyond
analysis that are traditionally used, PK/PD and QSP modeling the length of the fixed treatment duration. This is likely because,

CLINICAL PHARMACOLOGY & THERAPEUTICS | VOLUME 0 NUMBER 0 | Month 2024 19

unlike antibody-dependent cellular toxicity or antibody–drug con- risk prior to FIH studies using in silico sequence-based methods,
jugate mechanisms, TCEs are known to elicit deeper response that in vitro studies, and measuring pre-existing antibodies. The major-
is durable over a prolonged time. Fixed treatment duration strategy ity of the currently approved TCEs are B-cell targeting molecules
is promising in terms of patient compliance and minimizing any and thus have shown limited occurrence of ADAs in the clinic or
long-term toxicities. However, mechanistic rationale to accurately their impact on PK, efficacy, and safety. As TCEs successfully ex-
define the fixed treatment duration is yet to evolve, and additional pand to other targets and disease areas, ADAs may become a key
PK/PD-based investigations and clinical studies would be needed developability consideration.
to implement fixed treatment duration across TCEs. In addition,
fixed treatment duration strategies are likely to vary depending on Target engagement and biomarkers
the disease characteristics. RO and soluble targets are commonly evaluated as part of the
target engagement for TCEs. Characterization of soluble targets
DDI evaluation such as sBCMA has also proven to inform tumor burden and the
Drug–drug interaction (DDI) evaluations for TCEs generally impact of soluble target engagement on PK and dose.56 As sum-
follow the same principles as those for monoclonal antibodies, marized by Saber et al., RO has not been a relevant estimator of
except for the need to assess the effects of transiently elevated cy- clinical activity for TCEs. Therefore, RO data alone may not drive
tokine levels following the administration of TCEs. The elevation clinical decisions during the FIH study. However, RO data can
of cytokines, particularly IL6, has been shown to suppress cyto- be used with the observed clinical PK, efficacy, and safety data to
chrome P450 (CYP450) enzymes in humans. Given the transient derive mechanistic insights.
nature of cytokine elevation, the assessment of cytokine effects on Biomarker evaluation typically includes assessment of target
CYP450 is generally done with the help of a physiologically-based burden, cytokines, and T-cell dynamics. Cytokine measurements
pharmacokinetic (PBPK) model, as first implemented for blinatu- are useful to evaluate the occurrence of CRS, and to assess drug–
momab (Table 4).55 drug interactions (DDI) potential to inform the use of concomi-
In the case of teclistamab, a PBPK model was used to estimate tant medicines. T-cell dynamics are important to assess the MoA
the exposure of substrates of CYP1A2 (caffeine), CYP2C9 (S- and to derive PK/PD relationships.
warfarin), CYP2C19 (omeprazole), and CYP3A4/3A5 (mid-
azolam, cyclosporine, and simvastatin) with and without the FUTURE CONSIDERATIONS AND CONCLUSIONS
administration of teclistamab. The mean observed IL-6 profile was TCEs that specifically bind to tumor surface antigens and the
estimated to cause no clinically significant interaction with any of CD3ε chain of the TCR are a clinically validated modality for
the substrates (i.e., <20% change in substrate AUC). When the ef- the treatment of cancer, particularly hematological tumors. In the
fects were assessed for a patient with the highest IL6 levels, CYP future, TCEs will continue to evolve, driven by the transition of
activity was estimated to return to at least 80% of baseline within hemotologic cancer therapeutics toward solid tumor indications,
~7–8 days of the first treatment dose.51 Although there are uncer- which may require higher potency due to the restrictions of bio-
tainties around the clinical relevance of the effects of cytokines on distribution into solid tumors and a more immune suppressive
CYP enzymes, PBPK-based assessments have been successfully tumor microenvironment. TCEs will also evolve to reduce safety
used by all approved TCEs to inform labeling. No dedicated co- concerns with the first-generation molecules, including on-target/
horts or clinical studies evaluating the drug interaction potential of off-tumor toxicities and CRS which are major challenges for the
TCEs have been conducted for any of the approved TCEs. clinical development of this modality.
The complexity of TCEs will continue to increase to address
BIOANALYTICAL AND BIOMARKER STRATEGIES TO both efficacy and safety concerns. For example, traditional bispe-
SUPPORT CLINICAL PHARMACOLOGY ASSESSMENTS cific TCEs are being expanded to trispecific TCEs to include co-
Bioanalytical (pharmacokinetics and immunogenicity) stimulatory signals to foster efficient T-cell activation and cytokine
Pharmacokinetics and immunogenicity are key bioanalytical read- conjugation to stimulate immune response in solid tumors. Pro-
outs that guide clinical development for TCEs. Pharmacokinetic drug TCEs with steric masks activated by tumor-specific proteases
measurements can be assessed as either free active molecule (e.g., are being developed to enhance safety and open up target space
intact molecule not bound to soluble fractions) or total drug (with for TCEs. TCEs are also being designed to target innate-like γδ T
both bound and unbound fractions). When the total drug is mea- cells which may improve safety and efficacy by targeting a specific
sured, PK data should be interpreted along with soluble target ki- T-cell subset that can be further expanded in patients. Furthermore,
netics to assess the effect of soluble target engagement and ADA substantial efforts are underway to develop next-generation im-
data. For immunogenicity evaluations, it is generally accepted mune cell engagers targeting NK cells or myeloid-derived cells for
to measure only the total anti-drug antibodies during the dose- orthogonal MoA. As complexity increases, challenges in clinical
escalation phase to guide clinical decisions. However, within the development highlighted in this article including clinical dose set-
registration studies, it is important to characterize the ADA to ting, trial design, immunogenicity mitigation, and patient selection
understand the neutralizing ability of the observed ADA and its will require careful navigation. Tools such as mechanistic model-
impact on PK, efficacy, and safety. ing, combined with biomarkers of efficacy, and precision-medicine
Given the intricate complexity in the molecular formats for approaches will be essential to facilitate the clinical development of
TCEs, it is generally recommended to assess immunogenicity next-generation TCE molecules.

20 VOLUME 0 NUMBER 0 | Month 2024 | www.cpt-journal.com

Through this white paper, we have provided our perspectives 10. Frances, N. et al. Novel in vivo and in vitro pharmacokinetic/
Pharmacodynamic-based human starting dose selection for
on FIH and clinical pharmacology strategies to support early Glofitamab. J Pharm Sci 111, 1208–1218 (2022).
clinical development of TCEs. The pharmacological assay set- 11. Chari, A. et al. Talquetamab, a T-cell-redirecting GPRC5D
ups play a critical role in driving the selection of optimal starting bispecific antibody for multiple myeloma. N Engl J Med 387,
dose. Clinical top dose and RP2D decisions should be supported 2232–2244 (2022).
12. Zhu, M. et al. Translational findings for odronextamab: from
by a thorough evaluation of clinical pharmacokinetics, efficacy, preclinical research to a first-in-human study in patients with
and safety evaluation across multiple cohorts in conjunction with CD20+ B-cell malignancies. Clin Transl Sci 15, 954–966 (2022).
MIDD approaches. Additionally, preclinically (in vitro and/or 13. Qi, T., Liao, X. & Cao, Y. Development of bispecific T cell
engagers: harnessing quantitative systems pharmacology. Trends
in vivo) derived exposure estimates for maximal activity or tar- Pharmacol Sci 44, 880–890 (2023).
get engagement can further inform the determination of RP2D. 14. Chen, X. et al. Mechanistic projection of first-in-human dose for
While our recommendations in the paper are aimed to provide a bispecific immunomodulatory P-cadherin LP-DART: an integrated
PK/PD modeling approach. Clin Pharmacol Ther 100, 232–241
general framework toward early clinical development, one must be (2016).
cautious while translating learnings to a different molecule, target, 15. Vrohlings, M. et al. 195 overcoming the dose-response prediction
and/or disease indication. As evident from the case studies, the ap- limitation from bench to clinic for T-cell engagers: using
proaches used for approved molecules do have some overlaps in the quantitative systems pharmacology (QSP) modelling in the
development of CDR404 for solid tumors. J Immunother Cancer
methodologies; however, they also vary in the MIDD approaches 11(Suppl 1), A223 (2023).
that supported dose expansion strategies and RP2D selections. In 16. Flowers, D. et al. A next generation mathematical model for the
addition, there are still several unknowns in the field related to the in vitro to clinical translation of T-cell engagers. J Pharmacokinet
Pharmacodyn 50, 215–227 (2023).
T cell, target, and disease biology that preclude a priori predict- 17. Lemarre, P. et al. Cytotoxic effects of bispecific T-cell engagers
ability of clinical outcomes. Therefore, we advise translational and in QSP models under the lends of cell encounter [Conference
clinical pharmacologists referring to this white paper to carefully Poster]. In American Conference of Pharmacometrics, ACoP14.
Maryland (2023).
build fit-for-purpose strategies to inform the development of their
18. Deng, R., Iyer, S., Theil, F.P., Mortensen, D.L., Fielder, P.J. &
novel TCEs by using our recommendations as a general framework. Prabhu, S. Projecting human pharmacokinetics of therapeutic
antibodies from nonclinical data: what have we learned? MAbs 3,
FUNDING 61–66 (2011).
No funding was received for this work. 19. Ling, J., Zhou, H., Jiao, Q. & Davis, H.M. Interspecies scaling of
therapeutic monoclonal antibodies: initial look. J Clin Pharmacol
49, 1382–1402 (2009).
CONFLICTS OF INTEREST
20. Oitate, M. et al. Prediction of human pharmacokinetics of
All authors declared no competing interests for this work.
therapeutic monoclonal antibodies from simple allometry of
monkey data. Drug Metab Pharmacokinet 26, 423–430 (2011).
© 2024 The Author(s). Clinical Pharmacology & Therapeutics 21. Avery, L.B. et al. Utility of a human FcRn transgenic mouse model
© 2024 American Society for Clinical Pharmacology and Therapeutics. in drug discovery for early assessment and prediction of human
pharmacokinetics of monoclonal antibodies. MAbs 8, 1064–1078
(2016).
1. Goebeler, M.E. & Bargou, R.C. T cell-engaging therapies – BiTEs 22. Betts, A. et al. Linear pharmacokinetic parameters for monoclonal
and beyond. Nat Rev Clin Oncol 17, 418–434 (2020). antibodies are similar within a species and across different
2. Tapia-Galisteo, A., Alvarez-Vallina, L. & Sanz, L. Bi-and trispecific pharmacological targets: a comparison between human,
immune cell engagers for immunotherapy of hematological cynomolgus monkey and hFcRn Tg32 transgenic mouse using a
malignancies. J Hematol Oncol 16, 83 (2023). population-modeling approach. MAbs 10, 751–764 (2018).
3. Paz-Ares, L. et al. Tarlatamab, a first-in-class DLL3-targeted 23. Saber, H., Thompson, M.D. & Leighton, J.K. Pharmacokinetic
bispecific T-cell engager, in recurrent small-cell lung cancer: an models for first-in-human dose selection of immune-activating
open-label, phase I study. J Clin Oncol 41, 2893–2903 (2023). products in oncology. Regul Toxicol Pharmacol 149, 105616
4. Kelly, W.K. et al. Xaluritamig, a STEAP1 × CD3 XmAb 2+1 immune (2024).
therapy for metastatic castration-resistant prostate cancer: 24. Optimizing Dosing in Oncology Drug Development [white paper].
results from dose exploration in a first-in-human study. Cancer Friends of Cancer Research Annual Meeting [Internet] (2021)
Discov 14, 76–89 (2024). <https://friendsofcancerresearch.org/wp- content/uploads/Optim
5. Saber, H., del Valle, P., Ricks, T.K. & Leighton, J.K. An FDA izing_ Dosing_in_Oncology_ Drug_ Development.pdf> [Accessed
oncology analysis of CD3 bispecific constructs and first-in-human 2024].
dose selection. Regul Toxicol Pharmacol 90, 144–152 (2017). 25. Shah, M., Rahman, A., Theoret, M. & Pazdur, R. How to Get the
6. Moreau, P., Girgis, S. & Goldberg, J.D. Teclistamab in relapsed or Dose Right [Conference Presentation]. In American Society of
refractory multiple myeloma. Reply N Engl J Med 387, 1722–1723 Clinical Oncology. Chicago (2022).
(2022). 26. Fourie Zirkelbach, J. et al. Improving dose-optimization processes
7. Jiang, X. et al. Development of a target cell-biologics-effector used in oncology drug development to minimize toxicity and
cell (TBE) complex-based cell killing model to characterize target maximize benefit to patients. J Clin Oncol 40, 3489–3500 (2022).
cell depletion by T cell redirecting bispecific agents. MAbs 10, 27. Optimizing the Dosage of Human Prescription Drugs and Biological
876–889 (2018). Products for the Treatment of Oncologic Diseases [Draft Guidance
8. Liao, M.Z. et al. Optimization of FIH (First-in-Human) Starting Dose for Industry] [Internet] (FDA Center for Drug Evaluation and
for an HLE-BiTE® (Half-life Extended Bispecific T-cell Engager Research, 2023) <https://w ww.fda.gov/regulatory-information/
Molecule for Patients with Gastric Cancer Using Translational search-fda-guidance- documents/optimizing- dosage-human-
PKPD Modeling. [Conference Poster]). In American Conference on prescription- drugs-and- biological-products-treatment- oncol
Pharmacometrics (ACoP) (2020). ogic- diseases>.
9. Krupka, C. et al. CD33 target validation and sustained depletion 28. Budde, E.L. et al. Subcutaneous Mosunetuzumab is active with
of AML blasts in long-term cultures by the bispecific T-cell- a manageable safety profile in patients (pts) with relapsed/
engaging antibody AMG 330. Blood 123, 356–365 (2014). refractory (R/R) B-cell non-Hodgkin lymphomas (B-NHLs): updated

CLINICAL PHARMACOLOGY & THERAPEUTICS | VOLUME 0 NUMBER 0 | Month 2024 21

results from a phase I/II study. Blood 140(Supplement 1), of subcutaneous Epcoritamab to mitigate CRS risk using repeated
3753–3755 (2022). time-to-event modeling [abstract]. In Proceedings of the American
29. Lim, E.A. et al. Phase 1 study of safety and preliminary clinical Association for Cancer Research Annual Meeting 2023; Part 1
activity of JNJ-63898081, a PSMA and CD3 bispecific antibody, (Regular and Invited Abstracts); 2023 Apr 14–19; Orlando, FL and
for metastatic castration-resistant prostate cancer. Clin Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract
Genitourin Cancer 21, 366–375 (2023). nr 2798 (2023).
30. Falchi, L., Vardhana, S.A. & Salles, G.A. Bispecific antibodies 45. Li, T. et al. Semimechanistic physiologically-based
for the treatment of B-cell lymphoma: promises, unknowns, and pharmacokinetic/Pharmacodynamic model informing Epcoritamab
opportunities. Blood 141, 467–480 (2023). dose selection for patients with B-cell lymphomas. Clin Pharmacol
31. Hendrikx, J. et al. Fixed dosing of monoclonal antibodies in Ther 112, 1108–1119 (2022).
oncology. Oncologist 22, 1212–1221 (2017). 46. Yuwen, H. et al. ATG-101 is a tetravalent PD-L1x4-1BB bispecific
32. Van De Vyver, A. et al. A novel approach for quantifying the antibody that stimulates anti-tumor immunity through PD-L1
pharmacological activity of T-cell engagers utilizing in vitro time blockade and PD-L1-directed 4-1BB activation. Cancer Res 84,
course experiments and streamlined data analysis. AAPS J 24, 7 1680–1698 (2024).
(2021). 47. Ahn, M.J. et al. Tarlatamab for patients with previously treated
33. Shimabukuro-Vornhagen, A. et al. Cytokine release syndrome. J small-cell lung cancer. N Engl J Med 389, 2063–2075 (2023).
Immunother Cancer 6, 56 (2018). 48. Susilo, M.E. et al. Systems-based digital twins to help
34. Salvaris, R., Ong, J. & Gregory, G.P. Bispecific antibodies: a review characterize clinical dose-response and propose predictive
of development, clinical efficacy and toxicity in B-cell lymphomas. biomarkers in a phase I study of bispecific antibody,
J Pers Med 11, 355 (2021). mosunetuzumab. Clin Transl Sci 16, 1134–1148 (2023).
35. Mohamed Elmeliegy, A.V. et al. Dose optimization to mitigate the 49. Liu, C., Zhou, J., Kudlacek, S., Qi, T., Dunlap, T. & Cao, Y.
risk of CRS with Elranatamab in multiple myeloma. Blood 140, Population dynamics of immunological synapse formation
7174–7175 (2022). induced by bispecific T cell engagers predict clinical
36. Ball, K., Dovedi, S.J., Vajjah, P. & Phipps, A. Strategies for clinical pharmacodynamics and treatment resistance. eLife 12, 12
dose optimization of T cell-engaging therapies in oncology. MAbs (2023).
15, 2181016 (2023). 50. Oncology Dosing Tool Kit [Guidance for Industry] [Internet]
37. Chen, X., Kamperschroer, C., Wong, G. & Xuan, D. A modeling (Oncology Center of Excellence, US FDA, 2023).
framework to characterize cytokine release upon T-cell-engaging 51. Johnson & Johnson Innovative Medicine. TECVAYLI™ (teclistamab-
bispecific antibody treatment: methodology and opportunities. cqyv) [NDA/BLA Multi-disciplinary Review and Evaluation; BLA
Clin Transl Sci 12, 600–608 (2019). 761291]. US Food and Drug Administration website. October
38. Hosseini, I. et al. Mitigating the risk of cytokine release 2022 <https://w ww.accessdata.fda.gov/drugsatfda_docs/nda/
syndrome in a phase I trial of CD20/CD3 bispecific antibody 2022/761291Orig1s000MultidisciplineR.pdf>. Accessed Feb
mosunetuzumab in NHL: impact of translational system modeling. 2024.
NPJ Syst Biol Appl 6, 28 (2020). 52. Amgen Inc. BLINCYTO® (Blinatumomab) [Package Insert] (US Food
39. Pfizer Inc. ELREXFIOTM (elranatamab-bcmm). [package insert] (US and Drug Administration, 2023) [Internet] <https://w ww.acces
Food and Drug Administration, 2024) <https://w ww.accessdata. sdata.fda.gov/drugsatfda_docs/label/2024/125557Orig1s028
fda.gov/drugsatfda_docs/label/2023/761345s000lbl.pdf>. Correctedlbl.pdf>. Accessed Feb 2024.
Accessed Feb 2024. 53. Genentech, Inc. LUNSUMIO™ (Mosunetuzumab-A xgb) [Package
40. Van De Vyver, A.J. et al. Cytokine release syndrome by T-cell- Insert] (US Food and Drug Administration, 2022) <https://w ww.
redirecting therapies: can we predict and modulate patient risk? accessdata.fda.gov/drugsatfda_docs/label/2022/761263s000
Clin Cancer Res 27, 6083–6094 (2021). lbl.pdf>. Accessed Feb 2024.
41. Budde, L.E. et al. Single-agent Mosunetuzumab shows durable 54. Genentech, Inc. COLUMVI (Glofitamab-Gxbm) [Package Insert] (US
complete responses in patients with relapsed or refractory B-cell Food and Drug Administration, 2023) <https://w ww.accessdata.
lymphomas: phase I dose-escalation study. J Clin Oncol 40, fda.gov/drugsatfda_docs/label/2023/761309s000lbl.pdf>.
481–491 (2022). Accessed Feb 2024.
42. Li, T. et al. Identification of optimal dosing regimen for 55. Xu, Y., Hijazi, Y., Wolf, A., Wu, B., Sun, Y.N. & Zhu, M.
subcutaneous Epcoritamab in relapsed or refractory B-cell non- Physiologically based pharmacokinetic model to assess the
Hodgkin lymphoma. Blood 142(Suppl 1), 3135 (2023). influence of Blinatumomab-mediated cytokine elevations on
43. Girgis, S. et al. Translational modeling predicts efficacious cytochrome P450 enzyme activity. CPT Pharmacometrics Syst
therapeutic dosing range of Teclistamab for multiple myeloma. Pharmacol 4, 507–515 (2015).
Target Oncol 17, 433–439 (2022). 56. Girgis, S. et al. Effects of teclistamab and talquetamab on
44. Tommy Li, D.P., Thalhauser, C., Parikh, A., Gupta, M., Sacchi, M. soluble BCMA levels in patients with relapsed/refractory multiple
& Xu, S. Simplifying selection and optimization of step-up dosing myeloma. Blood Adv 7, 644–648 (2023).

22 VOLUME 0 NUMBER 0 | Month 2024 | www.cpt-journal.com

Strategies For Clinical Dose Optimization of T Cell-Engaging Therapies in Oncology
No ratings yet
Strategies For Clinical Dose Optimization of T Cell-Engaging Therapies in Oncology
18 pages
ESMO 2022 Reck Presentation - 0
No ratings yet
ESMO 2022 Reck Presentation - 0
28 pages
7070 - Downloadable Slides - ONC 2024 Heme Bispecifics
No ratings yet
7070 - Downloadable Slides - ONC 2024 Heme Bispecifics
52 pages
Ca Immunotherapy Beyond PDF
No ratings yet
Ca Immunotherapy Beyond PDF
25 pages
CARs Versus BiTEs - A Comparison Between T Cell-Redirection Strategies For Cancer Treatment
No ratings yet
CARs Versus BiTEs - A Comparison Between T Cell-Redirection Strategies For Cancer Treatment
12 pages
924 Full PDF
No ratings yet
924 Full PDF
12 pages
Glofitamab CD20-TCB Bispecific Antibody
No ratings yet
Glofitamab CD20-TCB Bispecific Antibody
12 pages
Batlevi Rev Immunotherapy
No ratings yet
Batlevi Rev Immunotherapy
16 pages
CAR T Therapy Beyond Cancer: The Evolution of A Living Drug
No ratings yet
CAR T Therapy Beyond Cancer: The Evolution of A Living Drug
14 pages
Current Immunotherapeutic Strategies in Cancer Matthias Theobald Download
No ratings yet
Current Immunotherapeutic Strategies in Cancer Matthias Theobald Download
93 pages
Bispecific Antibody Innovations
No ratings yet
Bispecific Antibody Innovations
27 pages
F6etvc9ye50s BsAbinLymphomaSlideHandout
No ratings yet
F6etvc9ye50s BsAbinLymphomaSlideHandout
46 pages
Poster - Cayatte C - Preclinical Dev of A AZD5492, CD8guided TCE For BNHL - EHA2024
No ratings yet
Poster - Cayatte C - Preclinical Dev of A AZD5492, CD8guided TCE For BNHL - EHA2024
1 page
Bite For Cancer Immunotherapy - Expert Opin Biol Ther 2015
No ratings yet
Bite For Cancer Immunotherapy - Expert Opin Biol Ther 2015
8 pages
Cancer Modelling
100% (1)
Cancer Modelling
15 pages
2023-R-Current Approaches To Develop Off-The-Shelf CAR-T For Cancer Treatment
No ratings yet
2023-R-Current Approaches To Develop Off-The-Shelf CAR-T For Cancer Treatment
17 pages
BTB8 Cancer Therapy
No ratings yet
BTB8 Cancer Therapy
48 pages
Current Immunotherapeutic Strategies in Cancer Matthias Theobald Instant Download
No ratings yet
Current Immunotherapeutic Strategies in Cancer Matthias Theobald Instant Download
122 pages
7 Cancer Immunotherapy Etc-2020
No ratings yet
7 Cancer Immunotherapy Etc-2020
32 pages
TP 4 Immunotherapeutics
No ratings yet
TP 4 Immunotherapeutics
37 pages
Management and Toxicity of Bi-Specifics: Luciano Costa, MD University of Alabama at Birmingham Birmingham, Alabama
100% (1)
Management and Toxicity of Bi-Specifics: Luciano Costa, MD University of Alabama at Birmingham Birmingham, Alabama
33 pages
2022-T-Cell Immunotherapy and Cardiovascular Disease
No ratings yet
2022-T-Cell Immunotherapy and Cardiovascular Disease
12 pages
Emerging New Therapeutic Antibody Derivatives For Cancer Treatment
No ratings yet
Emerging New Therapeutic Antibody Derivatives For Cancer Treatment
28 pages
TCR Cancer Therapeutics Overview
No ratings yet
TCR Cancer Therapeutics Overview
48 pages
2023 CAR T Cell Therapy - Remedies of Current Challenges in Design, Injection, Infiltration and Working
No ratings yet
2023 CAR T Cell Therapy - Remedies of Current Challenges in Design, Injection, Infiltration and Working
10 pages
Allogenic CAR-T Cells
No ratings yet
Allogenic CAR-T Cells
17 pages
N Weisser PEGS Europe2024
No ratings yet
N Weisser PEGS Europe2024
26 pages
CAR T Cells For Solid Tumors Developments To Watch in 2023
No ratings yet
CAR T Cells For Solid Tumors Developments To Watch in 2023
6 pages
Vaccines 11 00835
No ratings yet
Vaccines 11 00835
19 pages
Combination Therapy With T Cell Engager and PD-L1 Blockade Enhances The Antitumor Potency of T Cells As Predicted by A QSP Model
No ratings yet
Combination Therapy With T Cell Engager and PD-L1 Blockade Enhances The Antitumor Potency of T Cells As Predicted by A QSP Model
11 pages
Cancer Ii - 230522 - 095514
No ratings yet
Cancer Ii - 230522 - 095514
30 pages
Fukumura 2018
No ratings yet
Fukumura 2018
16 pages
Yoon 2018
No ratings yet
Yoon 2018
16 pages
CAR T Cell Therapy
No ratings yet
CAR T Cell Therapy
44 pages
Over 19 Million History of Cancer
No ratings yet
Over 19 Million History of Cancer
14 pages
Mechanisms of Therapeutic Antitumor Monoclonal Antibodies: Cancer Research - Review
No ratings yet
Mechanisms of Therapeutic Antitumor Monoclonal Antibodies: Cancer Research - Review
11 pages
CART Cell Therapy - A Versatile Living Drug
No ratings yet
CART Cell Therapy - A Versatile Living Drug
22 pages
Molecular Oncology - 2015 - Whilding - CAR T Cell Immunotherapy The Path From The by Road To The Freeway
No ratings yet
Molecular Oncology - 2015 - Whilding - CAR T Cell Immunotherapy The Path From The by Road To The Freeway
25 pages
BMJ 2021 068956.full
No ratings yet
BMJ 2021 068956.full
14 pages
64 - Online PDF 2
No ratings yet
64 - Online PDF 2
39 pages
Cancer DR 44-44
No ratings yet
Cancer DR 44-44
1 page
CAR T-Cell Therapy For B-Cell Lymphoma (Artigo - Leitura Vírginia)
No ratings yet
CAR T-Cell Therapy For B-Cell Lymphoma (Artigo - Leitura Vírginia)
12 pages
Immun Therapy
No ratings yet
Immun Therapy
2 pages
Pi Is 1535610820303743
No ratings yet
Pi Is 1535610820303743
19 pages
CAR Race To Cancer Immunotherapy: From CAR T, CAR NK To CAR Macrophage Therapy
No ratings yet
CAR Race To Cancer Immunotherapy: From CAR T, CAR NK To CAR Macrophage Therapy
21 pages
Neoantigen Cancer Vaccines: A New Star On The Horizon
No ratings yet
Neoantigen Cancer Vaccines: A New Star On The Horizon
38 pages
CART Project
No ratings yet
CART Project
11 pages
Targeted Chemotherapy Toxicity
No ratings yet
Targeted Chemotherapy Toxicity
41 pages
CAR T Therapy Beyond Cancer: The Evolution of A Living Drug: Perspective
No ratings yet
CAR T Therapy Beyond Cancer: The Evolution of A Living Drug: Perspective
9 pages
Advances in Car-T Cell Therapy Research
No ratings yet
Advances in Car-T Cell Therapy Research
27 pages
CAR T-Cell Therapy AJR 2021
No ratings yet
CAR T-Cell Therapy AJR 2021
14 pages
Cancers
No ratings yet
Cancers
22 pages
Advances in Cancer Immunotherapy 2019 - Latest Trends
No ratings yet
Advances in Cancer Immunotherapy 2019 - Latest Trends
21 pages
Cancer Vaccines From Research To Clinical Practice - 1st Edition Scribd Full Download
No ratings yet
Cancer Vaccines From Research To Clinical Practice - 1st Edition Scribd Full Download
17 pages
CAR T Cells: Solid Tumor Challenges
No ratings yet
CAR T Cells: Solid Tumor Challenges
9 pages
DR Francesca Lim
No ratings yet
DR Francesca Lim
37 pages
Spontaneous Intracerebral Hemorrhage
No ratings yet
Spontaneous Intracerebral Hemorrhage
37 pages
SF 1214
No ratings yet
SF 1214
4 pages
Stress 1
No ratings yet
Stress 1
36 pages
Work Ethic 1
No ratings yet
Work Ethic 1
18 pages
High Strength MAG Welding Wire Guide
No ratings yet
High Strength MAG Welding Wire Guide
1 page
Tools and Equipments in Food Preservation
No ratings yet
Tools and Equipments in Food Preservation
4 pages
Novice Program 1 Week 11 12 Gym Gear
No ratings yet
Novice Program 1 Week 11 12 Gym Gear
5 pages
B Finances Ss
No ratings yet
B Finances Ss
15 pages
Naturally Enhanced Eggs As A Source of Vitamin D
No ratings yet
Naturally Enhanced Eggs As A Source of Vitamin D
39 pages
Standardization of Methods of Analysis For Heavy Metals in Sediments
No ratings yet
Standardization of Methods of Analysis For Heavy Metals in Sediments
7 pages
METALS AND ALLOYS-4 - .Edited
No ratings yet
METALS AND ALLOYS-4 - .Edited
6 pages
Cooking Me Softly
No ratings yet
Cooking Me Softly
2 pages
Scopus Database Journal
No ratings yet
Scopus Database Journal
6 pages
Non Surgical Periodontal Therapy
No ratings yet
Non Surgical Periodontal Therapy
14 pages
Τeacher's Handbook - Preparing Candidates for the LRN Speaking Exams
100% (4)
Τeacher's Handbook - Preparing Candidates for the LRN Speaking Exams
104 pages
Drilling Engineering Competencies
100% (4)
Drilling Engineering Competencies
3 pages
A Song of Black and
No ratings yet
A Song of Black and
200 pages
SDS For Polyurethane Grout
No ratings yet
SDS For Polyurethane Grout
6 pages
Grad
No ratings yet
Grad
132 pages
The Tender Land by Kathleen Finneran - Discussion Questions
No ratings yet
The Tender Land by Kathleen Finneran - Discussion Questions
5 pages
Engineering Alloys (307) Lecture 7 Titanium Alloys I: David Dye
No ratings yet
Engineering Alloys (307) Lecture 7 Titanium Alloys I: David Dye
15 pages
Item Wise Details of Vendor Registration With Registration No. and Date
No ratings yet
Item Wise Details of Vendor Registration With Registration No. and Date
20 pages
Malakand University Scholarship Evaluation
No ratings yet
Malakand University Scholarship Evaluation
29 pages
FGIS Registered Grain Exporters Directory
33% (3)
FGIS Registered Grain Exporters Directory
2 pages
CRISP-DM & Business Understanding
No ratings yet
CRISP-DM & Business Understanding
5 pages
Dogville
No ratings yet
Dogville
2 pages
Essentials of Psychiatric Mental Health Nursing Concepts of Care in Evidence Based Practice 5th Edition Mary Townsend
100% (18)
Essentials of Psychiatric Mental Health Nursing Concepts of Care in Evidence Based Practice 5th Edition Mary Townsend
84 pages
Romeo & Juliet: A Student's Guide
No ratings yet
Romeo & Juliet: A Student's Guide
1 page
Q1 DLL Pe7 Week2
No ratings yet
Q1 DLL Pe7 Week2
9 pages
Serotonin
No ratings yet
Serotonin
1 page

Industry Perspective On First-In-Human and Clinical Pharmacology Strategies To Support Clinical Development of T-Cell Engaging Bispecific Antibodies For Cancer Therapy

Uploaded by

Industry Perspective On First-In-Human and Clinical Pharmacology Strategies To Support Clinical Development of T-Cell Engaging Bispecific Antibodies For Cancer Therapy

Uploaded by

WHITE PAPER

Industry Perspective on First-­in-­Human and

CLINICAL PHARMACOLOGY & THERAPEUTICS | VOLUME 0 NUMBER 0 | Month 2024 1

2 VOLUME 0 NUMBER 0 | Month 2024 | www.cpt-journal.com

CLINICAL PHARMACOLOGY & THERAPEUTICS | VOLUME 0 NUMBER 0 | Month 2024 3

4 VOLUME 0 NUMBER 0 | Month 2024 | www.cpt-journal.com

CLINICAL PHARMACOLOGY & THERAPEUTICS | VOLUME 0 NUMBER 0 | Month 2024 5

6 VOLUME 0 NUMBER 0 | Month 2024 | www.cpt-journal.com

CLINICAL PHARMACOLOGY & THERAPEUTICS | VOLUME 0 NUMBER 0 | Month 2024 7

8 VOLUME 0 NUMBER 0 | Month 2024 | www.cpt-journal.com

CLINICAL PHARMACOLOGY & THERAPEUTICS | VOLUME 0 NUMBER 0 | Month 2024

VOLUME 0 NUMBER 0 | Month 2024 | www.cpt-journal.com

CLINICAL PHARMACOLOGY & THERAPEUTICS | VOLUME 0 NUMBER 0 | Month 2024

following switch to Q2W.

switch from QW to Q2W

patients were able to

indicated that trimer

to tumor ratio was

and treatment doses via SC route. On one hand, starting an FIH

clinical PK and activity

tested positive for

phase. Additionally, for targets expressed in the skin and/or drug

of static effective concentrations in patients. On the other hand,

relatively higher Cmax-­to-­Ctrough ratio following IV administration

may allow sufficient time for T cells to disengage from continuous

tinuous T-­cell activation or exhaustion on the efficacy and safety

ports remain inconclusive to draw a correlation between the rate

inconclusive due to the differences in PK between IV and SC and

When switching between IV and SC routes, one must account

one can also use population PK modeling to select an SC dose that

leads to similar exposure as the IV dose. However, such analysis

regardless of IV and SC, the maximum tolerated exposures (respec-

Additionally, it may be possible for a TCE to be developed for both

BW vs. flat dose

12 VOLUME 0 NUMBER 0 | Month 2024 | www.cpt-journal.com

CLINICAL PHARMACOLOGY & THERAPEUTICS | VOLUME 0 NUMBER 0 | Month 2024 13

14 VOLUME 0 NUMBER 0 | Month 2024 | www.cpt-journal.com

CLINICAL PHARMACOLOGY & THERAPEUTICS | VOLUME 0 NUMBER 0 | Month 2024 15

16 VOLUME 0 NUMBER 0 | Month 2024 | www.cpt-journal.com

CLINICAL PHARMACOLOGY & THERAPEUTICS | VOLUME 0 NUMBER 0 | Month 2024 17

18 VOLUME 0 NUMBER 0 | Month 2024 | www.cpt-journal.com

CLINICAL PHARMACOLOGY & THERAPEUTICS | VOLUME 0 NUMBER 0 | Month 2024 19

20 VOLUME 0 NUMBER 0 | Month 2024 | www.cpt-journal.com

CLINICAL PHARMACOLOGY & THERAPEUTICS | VOLUME 0 NUMBER 0 | Month 2024 21

22 VOLUME 0 NUMBER 0 | Month 2024 | www.cpt-journal.com

You might also like

Industry Perspective on First-in-Human and

relatively higher Cmax-to-Ctrough ratio following IV administration

tinuous T-cell activation or exhaustion on the efficacy and safety