Volume 16, Issue 4 p. 694-703
ARTICLE
Open Access

A mechanistic PK/PD model to enable dose selection of the potent anti-tryptase antibody (MTPS9579A) in patients with moderate-to-severe asthma

Sharon M. Rymut

Sharon M. Rymut

Department of Clinical Pharmacology, Genentech Inc, South San Francisco, California, USA

Search for more papers by this author
Lindsay M. Henderson

Lindsay M. Henderson

Department of Clinical Pharmacology, Genentech Inc, South San Francisco, California, USA

Search for more papers by this author
Victor Poon

Victor Poon

Department of Clinical Pharmacology, Genentech Inc, South San Francisco, California, USA

Search for more papers by this author
Tracy L. Staton

Tracy L. Staton

Department of Ophthalmology, Metabolism, Neurology & Immunology Biomarker Development (OMNI-BD), Genentech Inc, South San Francisco, California, USA

Search for more papers by this author
Fang Cai

Fang Cai

Department of Ophthalmology, Metabolism, Neurology & Immunology Biomarker Development (OMNI-BD), Genentech Inc, South San Francisco, California, USA

Search for more papers by this author
Siddharth Sukumaran

Siddharth Sukumaran

Department of Preclinical and Translational PKPD, Genentech Inc, South San Francisco, California, USA

Search for more papers by this author
Horace Rhee

Horace Rhee

Early Clinical Development, Ophthalmology, Metabolism, Neurology Immunology (OMNI), Genentech Inc, South San Francisco, California, USA

Search for more papers by this author
Ryan Owen

Ryan Owen

Department of Clinical Pharmacology, Genentech Inc, South San Francisco, California, USA

Search for more papers by this author
Saroja Ramanujan

Saroja Ramanujan

Department of Preclinical and Translational PKPD, Genentech Inc, South San Francisco, California, USA

Search for more papers by this author
Kenta Yoshida

Corresponding Author

Kenta Yoshida

Department of Clinical Pharmacology, Genentech Inc, South San Francisco, California, USA

Correspondence

Kenta Yoshida, Genentech Inc., 1 DNA Way, South San Francisco, CA 94080, USA.

Email: [email protected]

Search for more papers by this author
First published: 08 February 2023

Sharon M. Rymut and Lindsay M. Henderson contributed equally to this work.

Abstract

Tryptase, a protease implicated in asthma pathology, is secreted from mast cells upon activation during an inflammatory allergic response. MTPS9579A is a novel monoclonal antibody that inhibits tryptase activity by irreversibly dissociating the active tetramer into inactive monomers. This study assessed the relationship between MTPS9579A concentrations in healthy subjects and tryptase levels in serum and nasal mucosal lining fluid from healthy subjects and patients with moderate-to-severe asthma. These data were used to develop a mechanistic pharmacokinetic/pharmacodynamic (PK/PD) model that quantitatively inter-relates MTPS9579A exposure and inhibition of active tryptase in the airway of patients with asthma. From initial estimates of airway tryptase levels and drug partitioning, the PK/PD model predicted almost complete neutralization of active tryptase in the airway of patients with asthma with MTPS9579A doses of 900 mg and greater, administered intravenously (i.v.) once every 4 weeks (q4w). Suppression of active tryptase during an asthma exacerbation event was also evaluated using the model by simulating the administration of MTPS9579A during a 100-fold increase in tryptase secretion in the local tissue. The PK/PD model predicted that 1800 mg MTPS9579A i.v. q4w results in 95.7% suppression of active tryptase at the steady-state trough concentration. Understanding how the exposure-response relationship of MTPS9579A in healthy subjects translates to patients with asthma is critical for future clinical studies assessing tryptase inhibition in the airway of patients with moderate-to-severe asthma.

Study Highlights

  • WHAT IS THE CURRENT KNOWLEDGE ON THE TOPIC?

MTPS9579A is a monoclonal antibody that inhibits tryptase activity by dissociating active tryptase tetramers into inactive monomers. In preclinical and clinical studies, MTPS9579A inhibited tryptase activity in the upper airway in a dose-dependent manner.

  • WHAT QUESTION DID THIS STUDY ADDRESS?

This study aimed to determine the relationship between MTPS9579A concentrations and tryptase levels by developing a mechanistic pharmacokinetic/pharmacodynamic (PK/PD) model that predicts tryptase inhibition in the airway of patients with moderate-to-severe asthma.

  • WHAT DOES THIS STUDY ADD TO OUR KNOWLEDGE?

The PK/PD model for MTPS9579A predicts almost full suppression of active tryptase in the airway in patients with moderate-to-severe asthma with doses of 900 mg and greater, administered intravenously once every 4 weeks. In the event of an asthma exacerbation, the model predicts 95.7% inhibition of active tryptase at the steady-state trough concentration with 1800 mg MTPS9579A dosed intravenously once every 4 weeks.

  • HOW MIGHT THIS CHANGE CLINICAL PHARMACOLOGY OR TRANSLATIONAL SCIENCE?

This study exemplifies the application of mechanistic PK/PD modeling to quantitatively analyze biomarkers in the target tissue and to inform the dose selection for future clinical studies.

INTRODUCTION

Asthma is a complex inflammatory airway disease that varies in molecular etiology, clinical presentation, and disease severity. Additional treatments are still needed for patients with severe asthma who do not achieve adequate symptom control with current therapies. In a real-world study of patients with asthma adherent to an inhaled corticosteroid and long-acting beta agonist treatment regimen, 81% of patients still experienced significant symptom burden and had an average of 1.3 exacerbations over the previous 12 months.1 Most current, successful therapies for asthma target mediators of type 2 inflammation such as interleukin (IL)-4, IL-5, and IL-13, which drive disease pathology.2, 3 However, these therapies provide inadequate symptom control for many patients whose asthma is driven by mechanisms that are independent of type 2 biology and who therefore require new therapies. Targeting tryptase, a protease generated by mast cells that promotes airway hyper-reactivity, remodeling, and inflammation,4 is a novel strategy to control type 2-independent asthma.

MTPS9579A is a fully humanized immunoglobulin (Ig) G4 monoclonal antibody that selectively inhibits tryptase activity by dissociating active tetramers into inactive monomers.5 A first-in-human phase I study showed MTPS9579A to be safe and tolerable in healthy subjects with favorable pharmacokinetics (PKs) at high doses.6 Nonlinear kinetics were observed at lower MTPS9579A concentrations. The phase I study also demonstrated that MTPS9579A distributes to the upper airway and neutralizes active tryptase in a dose-dependent manner in healthy subjects; serum total tryptase levels were elevated following administration of MTPS9579A.

There are several challenges in translating our current knowledge of MTPS9579A in nonclinical models and healthy subjects to patients with asthma.5 A diseased state involves multiple uncertainties that include physiological target levels, increases in target levels during an exacerbation event, amounts of drug present in the target tissue, and clearance mechanisms of the drug. For instance, mast cell degranulation during an asthma exacerbation event may be accompanied by elevated levels of tryptase in lung tissue, but the magnitude of this increase and its impact on overall tryptase activity and neutralization are not well understood.7-9

PK/pharmacodynamic (PK/PD) modeling enables the integration of molecular insights from nonclinical and healthy subject studies with knowledge of asthma pathophysiology to characterize a drug's exposure-response relationship. For this study, our goals were to quantitatively assess the relationship between MTPS9579A concentrations and tryptase levels in the airways of healthy subjects and to leverage this relationship to predict tryptase inhibition in the airways of patients with asthma. Initially, a preclinically developed PK/PD model for MTPS9579A, briefly described in a case report,10 helped guide molecule design and selection. We then expanded this mechanistic preclinical PK/PD model to incorporate PK data generated from healthy subjects, PD data from healthy subjects, and disease biomarker data from patients with moderate-to-severe asthma. We first predicted the extent of tryptase neutralization in patients with asthma at different MTPS9579A exposures. We also simulated a potential acute release of active tryptase in the airway during an asthma exacerbation and used our mechanistic PK/PD model to predict the dosing regimen necessary to achieve near-complete inhibition of active tryptase in this situation. For MTPS9579A, this PK/PD model can be utilized to mechanistically predict tryptase activity levels in the airway to inform dose selection in proof-of-concept clinical studies.

METHODS

Study design and subject population

Healthy subject PK/PD data were obtained from a phase I, randomized, observer-blinded, placebo-controlled study that investigated the safety, tolerability, PKs, and PDs of single and multiple ascending doses (MAD) of MTPS9579A.6 Briefly, the study was conducted in two parts: part A consisted of healthy subjects in seven single ascending-dose (SAD) cohorts receiving MTPS9579A either subcutaneously (s.c.; 30–300 mg) or intravenously (i.v.; 300–3600 mg); part B consisted of healthy subjects in five MAD cohorts receiving MTPS9579A (150–750 mg s.c. to 1350–3600 mg i.v.) once every 4 weeks (q4w). Reference data were collected at baseline from these subjects (n = 106). Samples obtained via nasosorption, a noninvasive, upper airway sampling method that uses a synthetic absorptive matrix to collect mucosal lining fluid samples from the nasal cavity,6, 11 were matched to serum samples also collected from the same subject.

Nasal mucosal lining fluid and matched serum samples in adults with moderate-to-severe asthma12 (n = 15; Table S1) requiring an inhaled corticosteroid and a second controller were provided by Dr. Dave Singh (UK Medicines Evaluation Unit [MEU]).

Biomarker sampling and outcome measures

For healthy subjects, nasosorption and serum samples were collected predose on day 1, and on days 2 (i.v. only), 5 (s.c. only), 15, and 29 for part A, and predose on day 1, and on days 57, 71, and 85 for part B.6 For patients with asthma, nasosorption and serum samples were collected at baseline, with repeat samples provided from six of 15 patients, 3 to 6 weeks after the baseline collection, resulting in a longitudinal dataset of 21 samples.

Total tryptase was measured in serum and nasosorption samples by enzyme-linked immunosorbent assay (lower limit of quantification [LLOQ]: 975 pg/ml) and Gyrolab Technology assay (LLOQ: 366 pg/ml; Gyros US, Inc.), respectively. An activity-based probe immunoassay on the Simoa platform was used to specifically measure active tryptase in nasosorption samples (LLOQ: 40 pg/ml).13 Monoclonal antibodies used for capture and detection for all assays were generated at Genentech, Inc. Samples with measured values below the LLOQ were imputed as LLOQ/2 for each assay. Nasosorption data were missing for some participants and/or timepoints because of insufficient sample volumes; no imputation was used for samples with missing data. To account for variability in sample volume absorbed onto the nasosorption devices during collection, tryptase levels in nasal lining fluid were normalized by the ratio of serum to airway urea individually.14 Median values across participants were determined to assess tryptase levels in the airway and serum.

Target-mediated drug disposition model for analysis of serum MTPS9579A and tryptase kinetics

A target-mediated drug disposition (TMDD) model was used to analyze PK/PD data from the phase I healthy subjects.6 MTPS9579A serum PK and tryptase levels in serum were obtained from all subjects who received at least one dose of MTPS9579A or placebo. In total, 835 serum PK and 937 serum PD samples from 106 subjects were included in the analysis. This TMDD model, which included two PK compartments and quasi-equilibrium (QE) approximation,15 was used to describe the observed serum concentrations and the PK/PD relationship (Figure 1; Text S1). QE approximation was selected because the conditions for simpler TMDD approximations (QE with a constant target amount and a free drug concentration > > target concentration for the Michaelis–Menten model) were not met. Body weight was included as a covariate in the TMDD model to account for its effects on both linear clearance and central volume of distribution. Parameter estimation was performed using nonlinear mixed-effects modeling with the Stochastic Approximation Expectation–Maximization algorithm in NONMEM (version 7.3., ICON Development Solutions).

Details are in the caption following the image
Overview of the tryptase and MTPS9579A mechanistic PK/PD model. The model consists of the following: (1) a systemic compartment for target-mediated drug disposition with tryptase monomers, (2) the distribution of MTPS9579A to target tissue, and (3) a target tissue compartment with the binding, interaction, or dissociation between MTPS9579A and the tryptase monomer or tetramer. See Methods section for details. i.v., intravenous; PK/PD, pharmacokinetic/pharmacodynamic; s.c., subcutaneous; t1/2, terminal half-life; t1/2,diss, terminal half-life for dissociation of active tetramer; t1/2,break, terminal half-life for tetramer disruption induced by MTPS9579A.

Mechanistic airway PK/PD model for simulation of local PK/PD

To simulate PK/PD relationships in the airway, we developed a mechanistic PK/PD model for MTPS9579A that combined kinetics of MTPS9579A and tryptase in serum (described with the TMDD model above), distribution of MTPS9579A and the MTPS9579A-tryptase complex to the target tissue, and MTPS9579A-tryptase interactions in the target tissue compartment (Figure 1). Kinetic parameters for the systemic compartment were fixed to those estimated in the previous step using serum PK/PD data. Other kinetic parameters were assigned based on known biological properties of tryptase and previous characterization of MTPS9579A (Table S2).

Distribution of MTPS9579A and the MTPS9579A-tryptase complex from circulation to airway interstitial fluid (ISF) was assumed to be determined by convective fluid flow combined with the vascular reflection coefficient representing the efficiency of antibody passing through vascular pores.16 As the influx and efflux fluid flow rates were assumed to be the same, the reflection coefficient determined the tissue ISF to serum concentration ratio (biodistribution coefficient). The biodistribution coefficient was assumed to be the same for MTPS9579A and the MTPS9579A-tryptase complex. The ISF fluid turnover rate constant for the lymphatic system was calculated as 0.2% of the lung plasma flow rate (182 L/h) divided by an ISF volume (0.3 L).17 The PK/PD model assumed similar target levels and antibody biodistribution across the upper and lower airway, based on the unified airway hypothesis.18

Tryptase in ISF was represented as being secreted in active tetramer form, with a physiologic irreversible dissociation rate constant (kdiss) of 33 day−1 (terminal half-life [t1/2,diss] of 30 min; Figure 1), resulting in formation of monomers with an elimination rate constant (kel) of 8.31 day−1 (half-life of 2 h).19 Binding interactions of MTPS9579A with tryptase β1 (both monomers and tetramers) were modeled using the following values: binding constant (kon): 7.62 × 105 (1/Ms); dissociation constant (koff): 3.72 × 10−5 (1/s); and equilibrium dissociation constant (KD): 4.88 × 10−11 (M), respectively.5 Additionally, a kdiss of 1000 day−1 (half-life [t1/2,break] of 1 min; Figure 1) was assumed for rapid, irreversible tetramer disruption induced by MTPS9579A (Table S2). A total (monomeric) tryptase concentration of 4 ng/ml in serum was based on measurements in healthy subjects5; an observed total tryptase concentration of 40 ng/ml in nasal lining fluid was based on patients with mild asthma (unpublished data). The PK/PD model (Figure 1, Text S2) and parameters used for simulations are summarized in Table S2. Simulations were performed with the mrgsolve package (version 0.11.1) in R 3.6.3 (https://cran.r-project.org/).

Ethics

For the phase I healthy subjects study, the clinical study protocol, any relevant associated documents, and informed consent forms (ICFs) were reviewed and approved by institutional review boards (IRBs; IRB Services, Aurora, Ontario, Canada; Advarra, Columbia, MD) before beginning study procedures, as previously described.6 For the patients with asthma study, the protocol, ICF, and any relevant associated documents were reviewed and approved by the South Manchester Local Research and Ethics Committee (Manchester, UK) before the trial started. For both studies, all participants provided written, informed consent.

RESULTS

Model-based analysis of phase I data and PK parameter estimations

Using the phase I clinical study results, we developed a mechanistic PK/PD model for MTPS9579A which includes the systemic circulation and the airway as the target tissue compartment. The QE TMDD model captured the observed serum PK data across a wide dose range, with both single (Figure 2a, Figure S1a) and multiple dose (Figure 2b, Figure S1a) administration (i.v. or s.c.) of MTPS9579A in healthy subjects. The model predicted effective target engagement, as demonstrated by elevated serum total tryptase levels (Figure 2c, Figure S1b). These data support our hypothesis that MTPS9579A has a high binding capacity for tryptase, resulting in TMDD, where serum total tryptase saturates lower MTPS9579A doses (30–300 mg s.c.). A critical concentration (Ccrit) for the onset of nonlinear PK for healthy subjects was estimated to be 120 nM (Ccrit = degradation rate constant [kdeg] * base/[clearance rate {CL}/volume of central compartment {Vc}]).20 The model-estimated PK parameters are provided in Table 1.

Details are in the caption following the image
Observed and simulated PK/PD profiles of MTPS9579A and tryptase in serum. Observed serum PK data following (a) SAD or (b) MADs of MTPS9579A and (c) corresponding serum total tryptase concentrations for the phase I healthy subject study. Observed data are represented by dots, with each color corresponding to the appropriate dose. Fitted lines represent the PK/PD model estimation. The solid black line (a, b) represents Ccrit, the critical drug concentration for linear clearance. i.v., intravenous; MAD, multiple ascending dose; PK/PD, pharmacokinetic/pharmacodynamic; SAD, single ascending dose; s.c., subcutaneous.
TABLE 1. Model-based human PK parameter estimates for MTPS9579A from the phase I healthy subject study
Parameters Unit Estimate RSE (%) Shrinkage (%)
First order absorption rate (ka) 1/day 0.239 7 -
Clearance (CL) L/day 0.128 4 -
Central volume of distribution (V2) L 3.33 4 -
Intercompartmental clearance (Q) L/day 0.408 7 -
Peripheral volume of distribution (V3) L 2.28 3 -
Subcutaneous bioavailability (Fsc) - 0.661 3 -
Thetarized proportional error for MTPS9579A - 0.103 4 -
Thetarized additive error for MTPS9579A - 2.72 23 -
Thetarized proportional error for total tryptase - 0.245 3 -
Equilibrium dissociation constant (KD) nM 0.0448 15 -
Baseline total tryptase (Base) nM 0.223 7 -
Total tryptase degradation rate constant (kdeg) 1/day 20.7 7 -
Clearance of MTPS9579A-total tryptase complex (CLint) L/day 0.398 5 -
Effect of weighta on CL (exponential model) - 0.82 23 -
Effect of weighta on V2 (exponential model) - 0.808 18 -
Interindividual variability (ω2)
k a - 0.214 26 40
CL - 0.0809 30 22
V 2 - 0.0387 26 23
Base - 0.197 17 8
k deg - 0.103 29 35
CLint - 0.135 25 22
  • Abbreviations: int, internalization; MAD, multiple ascending dose; PK, pharmacokinetic; RSE, relative standard error; SAD, single ascending dose.
  • a Weight range for subjects in the SAD stage was 49.9–113.3 kg and for the MAD stage was 47.1–97.7 kg.

MTPS9579A biodistribution coefficient predicted at 3% in the airway of healthy subjects

The mechanistic airway PK/PD model was then applied to evaluate the amount of MTPS9579A that partitions from the serum to the local site of action in the healthy subjects. Based on the unified airway disease hypothesis, which assumes that the lower and upper airway have similar MTPS9579A distribution characteristics, we tested three values for the antibody biodistribution coefficient (1%, 3%, or 10%; Figure S2) and evaluated the fit empirically by visual inspection. Model parameters estimated active tryptase levels in the airway relative to baseline over time for 30–3600 mg MTPS9579A administered either s.c. or i.v. (Figure 3a, solid lines). Observed levels of active tryptase from nasosorption samples collected in the phase I SAD study were superimposed onto the model-estimated active tryptase levels (Figure 3a, dotted lines) and were well captured assuming a 3% biodistribution coefficient (Figure S2a). Similar findings were observed after multiple dose administration and assuming a 3% biodistribution coefficient (Figure 3b, Figure S2b). Furthermore, the antibody biodistribution coefficient of 3% was in good agreement with the observed ratio of nasal lining fluid to serum MTPS9579A concentrations (Figure 4).

Details are in the caption following the image
Simulated and observed active tryptase suppression with the airway PK/PD model, assuming an upper airway interstitial tissue distribution for MTPS9579A of 3%. Observed active tryptase from healthy subjects (dots; dashed lines) and model-based upper airway active tryptase predictions (solid lines) following (a) SADs or (b) MADs of MTPS9579A. Dose levels are matched by color. Observed active tryptase data are represented as median percent of baseline. Subject samples below the limit of quantification in the baseline assessments were excluded. Data are non-normalized. i.v., intravenous; MAD, multiple ascending dose; PK/PD, pharmacokinetic/pharmacodynamic; q4w, every 4 weeks; SAD, single ascending dose; s.c., subcutaneous.
Details are in the caption following the image
Nasal mucosal lining fluid to serum PK concentration ratio. Ratio of urea-normalized nasal mucosal lining fluid to serum PK concentration over time following (a) SADs or (b) MADs of MTPS9579A in healthy subjects. Data points represent geometric mean and error bars represent geometric SD, with each color corresponding to the appropriate dose. i.v., intravenous; MAD, multiple ascending dose; PK, pharmacokinetic; q4w, every 4 weeks; SAD, single ascending dose; s.c., subcutaneous.

Baseline tryptase levels in airways are elevated in patients with asthma

To extrapolate MTPS9579A dosing regimens from healthy subjects to the target population, baseline airway tryptase levels from patients with moderate-to-severe asthma were analyzed (n = 15; mean age, 43 ± 13 years; 100% White; 27% women; Table S1). Median levels of urea-normalized active tryptase were elevated in patients with asthma compared with healthy subjects (4 ng/ml vs. 0.4 ng/ml; Figure 5a). Median levels of urea-normalized nasal total tryptase (140 ng/ml vs. 12 ng/ml in healthy subjects; Figure 5b) were also increased. Also integral for the PK/PD model was the observed elevation of the serum total tryptase levels, which was 1.5-fold higher in patients with asthma (10 ng/ml vs. 7 ng/ml in healthy subjects; Figure 5c).

Details are in the caption following the image
Baseline tryptase levels in patients with moderate-to-severe asthma compared with healthy subjects. (a) Nasal active tryptase, (b) nasal total tryptase, and (c) serum total tryptase. Boxplots show median, first, and third quartiles; whiskers extend to min/max values up to 1.5 times the interquartile range; dots represent individual subjects. For patients with asthma, the longitudinal dataset (n = 21) is displayed.

Forward predicting tryptase neutralization in the airway of patients with asthma

By incorporating the observed tryptase levels in patients with asthma into the tissue PD prediction model, we projected active tryptase levels (relative to baseline) over time for various MTPS9579A dose levels in the airway. The model was then used to predict the extent of target neutralization in patients with moderate-to-severe asthma. Assuming a 1.5-fold higher serum total tryptase concentration in patients with asthma at baseline compared with healthy subjects and an antibody biodistribution coefficient of 3%, the PK/PD model predicts that 900 mg MTPS9579A i.v. q4w results in 97.6% inhibition of active tryptase in the airway of patients with moderate-to-severe asthma (Table 2). Assuming that the target active tryptase suppression for a clinical study is greater than 95%, the model predicts that doses of 900 mg MTPS9579A i.v. q4w and greater will result in near-complete target suppression.

TABLE 2. Model-based predictions for percent inhibition of active tryptase by MTPS9579A in patients with moderate-to-severe asthma
Dose Tryptase multiplier
1× (no increase) 10× 100× 1000×
300 mg s.c. q4w 79.2 29.4 1.85 0.18
600 mg s.c. q4w 94.0 90.5 7.64 0.63
900 mg i.v. q4w 97.6 97.2 30.1 1.66
1800 mg i.v. q4w 98.9 98.8 95.7 3.87
3600 mg i.v. q4w 99.5 99.5 99.2 9.31
  • Note: Simulations used active tryptase levels ranging from baseline (1×; no increase) to tryptase increases from 10–1000-fold above baseline.

During an asthma exacerbation, mast cell degranulation may lead to a mass release of tryptase. Sensitivity analysis showed that tryptase neutralization is sensitive to the baseline levels of tryptase in the airway (Figure S3). Because the amount of tryptase released during an exacerbation is unknown, we projected additional scenarios to account for increases in the local tissue tryptase secretion rate ranging from 10–1000-fold. With a 100-fold increase, the model predicted that 1800 mg MTPS9579A i.v. q4w would result in 95.7% active tryptase suppression at the steady-state trough concentration. However, with a 1000-fold increase in tryptase secretion, the model predicted little active tryptase suppression (3.9%; Table 2).

DISCUSSION

Despite understanding the systemic characteristics of tryptase and MTPS9579A, little is known regarding active tryptase and MTPS9579A concentrations at the local site of action. For respiratory diseases, direct assessment of the lower airway requires more invasive procedures, such as bronchoscopy. Therefore, we developed a mechanistic PK/PD model based on a biomarker obtained from the less invasive nasosorption procedure to better understand the exposure-response relationship between MTPS9579A and active tryptase in healthy subjects. We then applied the model to patients with asthma to inform dose predictions for future proof-of-concept clinical studies.

The nonlinear kinetics of MTPS9579A and serum tryptase levels in healthy subjects were well described by a TMDD model. We hypothesized that TMDD was present based on MTPS9579A's relatively short half-life (7–11 days) at low doses (i.e., 30–300 mg s.c.) when compared to the expected IgG4 half-life (~21–28 days). The TMDD model captured this observation as demonstrated by the observed saturation of serum tryptase at higher doses (i.e., 900–3600 mg i.v.) and the longer observed half-lives at the 1800 and 3600 mg i.v. doses (30–35 days). The MTPS9579A-target complex had a two- to three-fold faster elimination than MTPS9579A alone, contributing to the nonlinear kinetics.

Previous reports have shown increases in baseline serum tryptase5 and bronchoalveolar lavage fluid tryptase21, 22 in patients with asthma. In addition to the elevated baseline serum tryptase results, our data also showed elevated baseline active and total tryptase levels in upper airway samples from patients with asthma compared with healthy subjects, furthering our understanding of tryptase biology. To bridge the gap between healthy subjects and patients with asthma, it was important to account for these elevated tryptase levels in simulating an asthmatic population. Our PK/PD model showed that the nonlinear PK threshold for MTPS9579A (Ccrit) would shift to higher concentrations for patients with asthma, as expected. Sensitivity analyses confirmed that the geometric mean of baseline tryptase levels and tryptase levels during an exacerbation were both important parameters to incorporate into the model. Although tryptase levels in the lungs are hypothesized to further increase during an asthma exacerbation, little is yet known about the magnitude of these increases. Because collecting nasosorption samples during an exacerbation would be difficult, doing so after nasal allergen challenge23 may provide surrogate data that overcomes this limitation and allows for model optimization.

To support appropriate dose selection such that a sufficient concentration of MTPS9579A reaches the site of action to neutralize the target, we expanded a preclinically developed PK/PD model10 to incorporate clinical data. We explored model parameters, their uncertainties, and differences between the patient population and the data used for model development. Based on a parameter sensitivity analysis, the biodistribution coefficient of free MTPS9579A to the target tissue and baseline tryptase levels in both serum and tissue were expected to have major impacts on the degree of active tryptase neutralization. However, other parameters, such as KD, lymphatic flow rate, tryptase degradation, and dissociation half-lives, were not expected to have major impacts within physiologically plausible ranges (data not shown). In our simulations, we accounted for elevated total tryptase levels in lung tissue during an asthma exacerbation by varying the acute increase in active tryptase, finding complete inhibition of active tryptase when we assumed a 100-fold increase in the lung tissue tryptase secretion rate. Furthermore, to develop an informative local exposure-response relationship, it was critical to estimate MTPS9579A biodistribution into the lower airway by testing a range of values for the biodistribution coefficient. An MTPS9579A biodistribution of 3% was well-supported by two key pieces of data: (1) the observed inhibition of nasosorption active tryptase was well-described by simulations varying the drug biodistribution to airway interstitial tissue (Figure S2) and (2) the observed nasal lining fluid-to-serum PK concentration ratio (Figure 4) across a wide range of s.c. and i.v. doses. We also assumed similar antibody tissue distribution between the upper and lower airway, based on the unified airway hypothesis.18 There was no obvious bias in the concentration ratio over time or at different dose levels, suggesting that tissue distribution of MTPS9579A was determined by nonspecific processes, such as convective transport.

The mechanistic PK/PD model will be a useful tool to inform dose selection of the anti-tryptase antibody MTPS9579A in patients with moderate-to-severe asthma in a proof-of-concept study because it enables the prediction of the MTPS9579A exposure and the magnitude of tryptase suppression in the target tissues. Further understanding of lower airway biodistribution and the impact of disease pathology on MTPS9579A localization and target levels will allow optimization of the PK/PD model and aid interpretation of subsequent clinical observations.

AUTHOR CONTRIBUTIONS

S.M.R., L.M.H., V.P., and K.Y. wrote the manuscript. S.M.R., T.L.S., F.C., S.S., H.R., R.O., S.R., and K.Y. designed the research. S.M.R., L.M.H., V.P., T.L.S., F.C., S.S., H.R., R.O., and K.Y. performed the research. S.M.R., L.M.H., V.P., T.L.S., F.C., S.S., H.R., R.O., S.R., and K.Y. analyzed the data. V.P. and K.Y. contributed new reagents/analytical tools.

ACKNOWLEDGMENTS

The authors thank Dr. Dave Singh at the UK Medicines Evaluation Unit (MEU) for providing nasosorption samples. We thank Prajna Banerjee for her contributions in managing the pharmacodynamic sampling and Viyia Sverkos in managing the pharmacokinetic sampling. Editing and writing assistance for this manuscript was provided by Deborah Solymar (Genentech, Inc.) and was funded by Genentech, Inc.

    FUNDING INFORMATION

    Genentech, Inc. supported this study.

    CONFLICT OF INTEREST STATEMENT

    S.M.R., L.M.H., V.P., T.L.S., F.C., S.S., H.R., R.O., S.R., and K.Y. are or were employees of Genentech, Inc., a member of the Roche Group, at the time this research was conducted and may hold Roche stock or stock options. F.C. is currently an employee and stockholder of Abbvie. S.S. is currently an employee of The Janssen Pharmaceutical Companies of Johnson & Johnson and is a stockholder of Johnson & Johnson.