Sometimes we come across an article we wish we had read earlier. This was the case with the publication by Zhao and colleagues[1]. The article, now more than 12 years old, was written in the ‘early’ days of biological drug development, after the first wave of mAbs had already been FDA reviewed and approved. Written by FDA pharmacologists, it distills their combined learnings from submissions, which in those days were predominantly humanized monoclonal antibodies.
The authors contrast the PK profiles of high MW biologicals, pointing out how their properties differ from NCEs. The PK behavior of biologicals is described for each stage of ADME, and also the specific impact of neutralizing ADAs. Receptor binding and processing via the FcRn and the Fcg family of receptors are unique features of mAbs affecting half-life significantly. Degradation in tissues is a factor even before mAbs reach the systemic circulation. Once locked onto a receptor, we need to be aware of target mediated drug disposition (TMDD). TMDD and receptor occupancy are dynamics that result in PK changes over time.
These concepts are probably known to pharmacologists working in biological drug development. This article is written by an authoritative group of pharmacologists at the FDA, who had the benefit of seeing and advising on many more projects than any of us in industry. Their condensed wisdom leads to a preferred PK/PD modeling approach of which the mathematics is provided in the article but we will skip in this blog.
Here some PK learnings and quotes:
- Absorption: the time to maximum serum concentration (Tmax) for mAbs ranges from 2 to 14 days.
This range is much larger than many would have predicted. - Absorption: a higher dose delivered at the injection site can saturate presystemic proteolytic degradation and may lead to an increase in bioavailability.
Makes sense, doesn’t it? - Distribution: Higher density of binding target in peripheral tissues and tighter binding to the target cell restrict mAb penetration deeper into tissues, subsequently resulting in limited distribution (i.e., binding-site barrier).
This barrier is important as it hinders penetration of a biological to reach the core of a solid tumor. - Metabolism/Excretion: IgA antibodies are mainly eliminated by biliary secretion.
In contrast, IgGs are usually broken down proteolytically in tissues unless recycled by the FcRn. - Metabolism/Excretion: IgGs engineered from different species have different affinities to human FcRn. As a result, ibritumomab, a murine IgG mAb, has a t1/2 of approximately 1 day in patients.
An interesting tidbit, but we deal almost exclusively nowadays with humanized mAbs, so this species difference becomes less relevant. - TMDD: Target-mediated clearance (CL) is saturable (i.e., capacity limited) because of the finite number of targets expressed on the cell surface.
This affects the PK/PD relationship which becomes non-linear. - ADA effect: Responders to infliximab treatment of ankylosing spondylitis had higher trough concentrations and lower incidence of HACAs (8%), whereas non-responders had lower trough concentrations and higher incidence of HACAs (36%).
Of course, this effect of ADAs has also been observed with adalimumab and other TNFα-blocking mAbs. It is the rationale for replacing mAbs with other biological formats like the nanos we described in our last blog. - Extrapolating HED: adequate nonclinical in vivo toxicological testing may be impossible because of the lack of cross-reactivity. Even for cross-reactive mAbs, the NOAEL obtained in test species may not be relevant to human testing.
So true – this has been a point of discussion with Regulators for many (if not all) biologicals. How can we do modeling if there is this species-barrier? Humanized animals have been created with the human receptor and the genetic make-up mimicking human immunology. We cannot adequately assess the pros and cons of this approach; suffice it to say that selection of the first dose in humans is meant to avoid a CRS a la TGN1412.
These factors make it difficult to study dose-finding in biologicals. There are regulatory guidelines for setting appropriate starting doses (MABEL) for FIH trials. This is only the first step. Ideally, one would like to establish the ‘right’ efficacious dose as early as possible. Safety/tolerability is rarely a guide for dose selection; most (but not all) mAbs have few off-target effects and are presumably ‘safe’ even at high doses. Of course, ISR can limit dose escalation, and also injection volume restricts SC administration to 2 mL for a single administration.
The non-linearity of the dose-response curve, the lack of AEs and dose-dependent safety signals make dose finding studies challenging. In addition, dosing frequency is important for patient acceptance as we are dealing with injectables. Less frequent dosing by spacing out injections is in the company’s interest but also in the interests of the patient. It is the lowest efficacious dose that many companies try to find, not the one with highest efficacy.
Nonetheless, dose escalation is easily testable, if a convenient efficacy read-out is available. IgG reduction in plasma is a reliable marker for efficacy with FcRn inhibitors. Batoclimab, an FcRn inhibitor under development by HarbourBiomed / Immunovent, provides a good example of dose-finding. High doses achieved 70% reduction of IgG and anti-AChR autoantibodies in gMG, the maximum efficacy of any FcRn biological. Unfortunately, undesirable effects on lipids and loss of blood albumin were observed, due to diminished recycling of albumin. These were on-target AEs and well-known class effects for the entire class of full-size IgG FcRn mAbs.
Here another 2 examples: UCB’s bimekizumab and MoonLake’s sonelokimab are both late-comers to the IL-17 mAb field vying for a slice of the market dominated by secukinumab/Cosentyx and several others. Existing mAbs have set a high hurdle for efficacy, with cure rates even for moderate/severe psoriasis in the 80% range. How is a new entry into this very competitive market to show differentiation in efficacy, a high hurdle to pass. Proving ‘superiority’ requires finding a highly effective dose that won’t create safety issues. It’s a tall order but can be done – it requires very careful data analysis and flexible prospective testing. We would really like to rely on PK/PD modeling for biologicals, the same way as we do routinely in antiinfective dose finding.[2]
In the case of batoclimab, a significant dose reduction became necessary. For bimekizumab, while best-in-class efficacy is still claimed by UCB with some justification, the chosen effective dose comes at the cost of deeper suppression of T-cell immunity and more frequent AEs due to candidiasis[3]. This is a manageable nuisance, but a nuisance nonetheless. It was a predictable AE as all IL-17 inhibitors have been associated with higher rates of candidiasis[4],[5]. For sonelokimab, the jury is still out, but some Phase 2 results indicate a propensity for mucocutaneous candidiasis as seen with the other IL-17A/F inhibitor, bimekizumab[6].
No discussion of PK/PD modeling should omit the TGN1412 disaster, a biological which was dosed too high in the first round of human testing. We just cannot rely entirely on preclinical testing and PK modeling for human dose selection in humans without accepting a residual risk. Keeping this risk to a minimum has to be the goal; it requires a humble attitude given our state of ignorance when testing novel biologic processes.
The Zhao article is a great primer on the PK issues encountered when developing biologicals. Highly recommended for study.
ABBREVIATIONS
ADA anti-drug antibody
ADME absorption, distribution, metabolism, excretion
CRS cytokine release syndrome
FcgR FC gamma receptor
FcRn neonatal Fc receptor
FIH first-in-human
gMG generalized myasthenia gravis
HACA human anti-chimeric antibodies
HED human equivalent dose
ISR injection site reaction
MABEL minimal anticipated biological effect level
PK/PD pharmacokinetics/pharmacodynamics
TMDD target-mediated drug deposition
REFERENCES
[1] Zhao L. Application of Pharmacokinetics–Pharmacodynamics/Clinical Response Modeling and Simulation for Biologics Drug Development. J Pharmaceutical Sciences, 101: 4367, 2012
[2] Many good references – please check publications by G. Drusano, J. Schentag, P. Ambrose, H. Derendorf, W. Craig and others
[3] Greer M. Bimekizumab-bkzx for the Treatment of Plaque Psoriasis: A Drug Review. Annals Pharmacother 59: 577, 2025
[4] Bilal H. Risk of candidiasis associated with interleukin-17 inhibitors: Implications and management. Mycology, 15: 30, 2023
[5] Davidson. Risk of candidiasis associated with interleukin-17 inhibitors: A real-world observational study of multiple independent sources. Lancet Regional Health – Europe, Volume 13, 100266, 2022
[6] Papp, K. IL17A/F nanobody sonelokimab in patients with plaque psoriasis: a multicentre, randomised, placebo-controlled, phase 2b study. Lancet, 397: 1564, 2021
