Elicit: Pharmacokinetics and Pharmacodynamics of Emicizumab in Hemophilia A

Pharmacokinetics and Pharmacodynamics of Emicizumab in Hemophilia A

Emicizumab exhibits predictable, dose-proportional pharmacokinetics with sustained plasma concentrations that produce FVIII-equivalent hemostatic activity and reduce bleeding rates by 94% through an exposure-response relationship plateauing above 30 µg/mL.

Abstract

Emicizumab demonstrates dose-proportional, predictable pharmacokinetics across diverse hemophilia A populations, with subcutaneous administration achieving sustained trough concentrations of 42-66 µg/mL during maintenance dosing across weekly (1.5 mg/kg), every-2-weeks (3 mg/kg), and every-4-weeks (6 mg/kg) regimens. The median elimination half-life is 23-30 days, with time to steady-state of 12 weeks with loading doses and 24 weeks without. Interindividual variability in trough concentrations is moderate (32%), influenced primarily by body weight, age, albumin levels, and rarely by neutralizing antibodies. Infants achieve higher concentrations (60-65 µg/mL) than older individuals with identical dosing, while bioavailability decreases after age 65. Pharmacodynamically, emicizumab maintains FVIII-equivalent activity of 17-25 IU/dL, thrombin generation >100 nM, and normalized aPTT without affecting FIX, FX, or coagulation activation markers.

Exposure-response modeling establishes that bleeding control plateaus above emicizumab concentrations of 30 µg/mL (IC50 3.58 µg/mL), explaining equivalent efficacy across dosing regimens that achieve mean concentrations of 53.5 µg/mL and produce a 94% reduction in annualized bleeding rates. Clinical studies demonstrate annualized bleeding rates for treated bleeds of 0.4-2.4, with 54-78% of patients achieving zero treated bleeds. The safety profile is favorable, with injection site reactions as the most common adverse event (16-22%), no thromboembolic events when used alone, and minimal immunogenicity. These pharmacokinetic and pharmacodynamic characteristics support body weight-based dosing without routine therapeutic drug monitoring, with potential for individualized dosing in patients maintaining concentrations substantially above the 30 µg/mL efficacy threshold.

Methods

We analyzed 10 sources from an initial pool of 200, using 8 screening criteria. Each paper was reviewed for 8 key aspects that mattered most to the research question. More on methods

Paper search

We performed a semantic search across over 138 million academic papers from the Elicit search engine, which includes all of Semantic Scholar and OpenAlex.

We ran this query: "Pharmacokinetics/pharmacodynamics of emicizumab in hemophilia A"

The search returned 200 total results from Elicit. We retrieved 200 papers most relevant to the query for screening.

Screening

We screened in sources based on their abstracts that met these criteria:

Population - Hemophilia A: Does the study include patients diagnosed with hemophilia A?
Intervention - Emicizumab: Does the study investigate emicizumab as the primary intervention?
Outcome - PK/PD Data: Does the study report pharmacokinetic data (e.g., plasma concentrations, clearance, half-life, bioavailability) and/or pharmacodynamic data (e.g., factor VIII mimetic activity, bleeding rates, coagulation parameters)?
Study Population - Human Subjects: Is the study conducted in human subjects (not animal or in vitro studies)?
Study Design: Is the study a clinical trial (Phase I, II, III, IV), observational study, case series, case report, systematic review, or meta-analysis?
Intervention Specificity: Does the study include emicizumab as an intervention (not focusing solely on other hemophilia treatments without emicizumab)?
Population Specificity: Does the study include hemophilia A participants (not exclusively patients with hemophilia B or other bleeding disorders without any hemophilia A participants)?
Data Relevance: Does the study report pharmacokinetic and/or pharmacodynamic data (not only clinical efficacy or safety outcomes without PK/PD data)?

Data extraction

Study Population: Extract patient demographics and clinical characteristics relevant to emicizumab pharmacokinetics/pharmacynamics in hemophilia A, including:

Age range and median/mean age
Body weight/BMI characteristics
Hemophilia A severity (mild, moderate, severe)
FVIII inhibitor status (presence/absence, titer levels if available)
Prior treatment history (treatment-naive, previously treated)
Geographic/ethnic population
Sample size for PK/PD analyses
Any other patient factors mentioned as potentially affecting emicizumab PK/PD

Dosing Regimen: Extract all details about emicizumab dosing and administration for PK/PD assessment, including:

Dose level(s) tested (mg/kg)
Dosing frequency (weekly, every 2 weeks, every 4 weeks)
Loading dose regimen if applicable
Maintenance dose regimen
Route of administration
Duration of treatment/observation
Any dose adjustments or individualization strategies
Sampling timepoints for PK/PD measurements

PK Parameters: Extract pharmacokinetic parameters and characteristics of emicizumab in hemophilia A patients, including:

Plasma/serum emicizumab concentrations (steady-state, peak, trough)
Clearance (CL/F)
Volume of distribution (V/F or Vss/F)
Half-life (t½)
Time to steady-state
Bioavailability or relative bioavailability
Dose-proportionality findings
PK variability (inter-patient, intra-patient)
Any population PK modeling results
Comparison across different dosing regimens if applicable

PD Outcomes: Extract pharmacodynamic outcomes and biomarkers for emicizumab in hemophilia A, including:

FVIII-equivalent activity levels (IU/dL or %)
Coagulation parameters (aPTT, thrombin generation, etc.)
Annualized bleeding rates (ABR) for treated bleeds, spontaneous bleeds, joint bleeds
Proportion of patients with zero bleeds
Time to first bleeding episode
Bleeding severity and location
Need for additional hemostatic treatment
Any other hemostatic efficacy measures
Methods used to assess each PD outcome

PK/PD Relationships: Extract any analyses of exposure-response or concentration-effect relationships for emicizumab in hemophilia A, including:

Correlations between emicizumab concentration and FVIII-equivalent activity
Exposure-efficacy relationships (concentration vs. bleeding rates)
Exposure-safety relationships
Population PK/PD modeling results
Threshold concentrations for efficacy
Model parameters (EC50, Emax, slope factors)
Goodness-of-fit assessments
Predictions or simulations of different dosing scenarios

Safety Profile: Extract safety and immunogenicity data relevant to emicizumab PK/PD in hemophilia A, including:

Injection site reactions
Thrombotic/thromboembolic events
Anti-drug antibodies (ADA) development and impact on PK/PD
Neutralizing antibodies
Other treatment-related adverse events
Laboratory safety parameters
Any safety findings that could affect dosing or PK/PD interpretation
Discontinuations due to safety concerns

Covariates: Extract factors identified as affecting emicizumab pharmacokinetics or pharmacodynamics in hemophilia A patients, including:

Age effects on PK/PD parameters
Body weight/size effects
Inhibitor status effects on efficacy
Prior treatment history effects
Genetic factors or ethnic differences
Concomitant medications effects
Disease-related factors affecting drug disposition
Any other identified sources of PK/PD variability
Magnitude and clinical significance of covariate effects

Study Methods: Extract methodological details relevant to PK/PD assessment of emicizumab in hemophilia A, including:

Study design and phase
PK sampling strategy and timepoints
Bioanalytical methods for emicizumab concentration measurement
Methods for measuring FVIII-equivalent activity and coagulation parameters
PK analysis methods (non-compartmental, population modeling software)
Statistical approaches for PK/PD analysis
Validation of bioanalytical and PD assays
Any limitations affecting PK/PD interpretation

Results

Characteristics of Included Studies

The review included 10 sources examining emicizumab pharmacokinetics and pharmacodynamics in hemophilia A, comprising clinical trial reports, a systematic review, and an exposure-response modeling study. The studies varied in design from Phase I dose-escalation trials to Phase III pivotal studies and represented diverse patient populations.

The populations spanned a wide age range from infants (9 days old) to adults (up to 77 years), including specialized studies in infants ≤12 months and Chinese pediatric patients. Most studies included patients with severe hemophilia A, with variable inhibitor status across studies. Prior treatment history varied, with several studies including previously untreated or minimally treated patients.

Dosing Regimens and Administration

Emicizumab was administered subcutaneously across all studies, with various loading and maintenance regimens tested to optimize pharmacokinetic profiles.

The standard loading regimen across recent studies was 3 mg/kg weekly for 4 weeks, designed to rapidly achieve therapeutic concentrations. Maintenance dosing flexibility allowed for once-weekly (1.5 mg/kg), every-2-weeks (3 mg/kg), or every-4-weeks (6 mg/kg) administration. Dose adjustments were permitted in some studies, with up-titration to 3 mg/kg weekly allowed for suboptimal efficacy after 6 months and individualized dose modifications implemented based on efficacy and safety committee review. In the HAVEN 7 study, one participant required dose up-titration from biweekly to weekly due to decreasing emicizumab levels.

Pharmacokinetic Parameters

Emicizumab demonstrated predictable, dose-proportional pharmacokinetics across diverse patient populations.

Plasma Concentrations

Trough concentrations increased during the loading phase, reaching mean values of 52.9-63.2 µg/mL by week 5, and were sustained at 42.1-66 µg/mL with maintenance dosing. The Chinese pediatric study reported somewhat lower median trough concentrations during maintenance (50.6 µg/mL). Dose-proportional increases in steady-state trough levels were observed across dose cohorts in the Phase I/II extension, with levels of 10.3, 29.9, and 120 mg/mL for 0.3, 1.0, and 3.0 mg/kg weekly dosing, respectively. Plasma concentrations increased in a dose-dependent manner across all dose escalation studies.

A notable finding was higher emicizumab concentrations in infants compared to older populations. In HAVEN 7, infant concentrations (60-65 µg/mL) were higher than those observed with the same dosing regimen in HAVEN 1-4 studies of older patients (46.7 ± 14.9 µg/mL) and HAVEN 2-3 studies (46-48 µg/mL).

Pharmacokinetic Model Parameters

The Chinese pediatric study provided detailed estimates using the published population PK model. Median clearance (CL/F) was 0.0637 L/day (range 0.0421-0.1578), volume of distribution (V/F) was 2.1543 L (range 1.2629-6.2103), and elimination half-life was 23.44 days (range 20.79-27.28). The earlier Phase I study reported a longer half-life of 4-5 weeks, while another source cited approximately 30 days. The systematic review noted that clearance was comparable to other therapeutic IgG4-based monoclonal antibodies (0.2-0.5 L/day), but emicizumab demonstrated increased distribution volumes (10 vs. 6 L) and a longer half-life. Time to steady-state was approximately 12 weeks with a loading dose and 24 weeks without loading.

Pharmacokinetic Variability

Interindividual variability in trough concentrations was moderate at 32% and remained consistent across various subgroups including FVIII inhibitor status, age group, and dosing interval. Total PK interindividual variability was approximately 60%, influenced by body weight, neutralizing antidrug antibodies, age, and albumin levels. Children showed slightly less variability than adults and adolescents. The long half-life was expected to result in minimal variation in peak/trough plasma concentrations, supporting stable pharmacokinetics with minimal peak-trough fluctuation.

Emicizumab demonstrated high subcutaneous bioavailability, with bioavailability decreasing with age, particularly after 65 years. The dose-linear pharmacokinetics supported body weight-based dosing, although individualized monitoring may allow for more cost-effective dosing.

Pharmacodynamic Outcomes

FVIII-Equivalent Activity and Coagulation Parameters

FVIII-like activity increased during the loading period to 21-25.2 IU/dL and was sustained at 17-23 IU/dL with maintenance dosing, remaining consistently above 20 U/dL. The Chinese pediatric study reported median FVIII-equivalent activity of 14.6 IU/dL during loading and 20.2 IU/dL during maintenance. These levels theoretically convert patients with severe hemophilia A to a mild disease phenotype.

Thrombin generation peak height increased during loading, reaching 100-115.2 nM by week 5-17, and was sustained above 100-116 nM thereafter. Activated partial thromboplastin time (aPTT) was normalized at subtherapeutic concentrations of emicizumab, shortened following the first dose, normalized by Week 3 in most participants, and remained short throughout studies.

Importantly, emicizumab did not significantly affect plasma concentrations of target antigens FIX and FX, nor did it affect fibrinogen, prothrombin time (international normalized ratio), D-dimer, or prothrombin fragment 1.2 levels, indicating no induction of consumptive coagulopathy or coagulation activation.

Bleeding Outcomes

Emicizumab prophylaxis demonstrated substantial efficacy in reducing bleeding rates across all studies. In the infant population (HAVEN 7), the model-based ABR for treated bleeds was 0.4 (95% CI: 0.23-0.65 in interim analysis and 0.30-0.63 in primary analysis), with ABR for all bleeds of 1.9-2.0 and treated joint bleeds of 0.0-0.1. No intracranial hemorrhage occurred in this vulnerable population. The proportion of patients achieving zero treated bleeds ranged from 54.5% to 77.8% in the HAVEN 7 analyses.

In the Phase I dose-escalation study, median ABRs decreased from baseline values of 18.3-32.5 to 0.0-4.4 across cohorts receiving 0.3-3.0 mg/kg weekly, with 71-73% of patients experiencing no bleeding. The long-term extension showed median ABRs remaining low at 0.0-1.4.

The HAVEN 4 study examining every-4-weeks dosing reported an ABR of 2.4 (95% CI: 1.4-4.3) for treated bleeds, with 56.1% of patients reporting no treated bleeds. The Chinese pediatric study achieved a median ABR of 0. Overall, the systematic review found that 70.8% of patients had no treated bleeds initially, increasing to 80.2% after 1 year of treatment.

The majority of bleeds were traumatic (87.9% in HAVEN 7), with breakthrough bleeding episodes successfully treated with episodic FVIII or bypassing agents when needed. The episodic use of clotting factors to control bleeding was reduced.

Exposure-Response Relationships

Several studies characterized the relationship between emicizumab concentrations and clinical outcomes, demonstrating clear concentration-dependent effects on both pharmacodynamic biomarkers and bleeding control.

Concentration-Biomarker Correlations

FVIII-like activity and thrombin generation peak height correlated with emicizumab concentrations. The Chinese pediatric study quantified this relationship, finding a positive correlation between emicizumab concentration and FVIII-equivalent activity by linear regression (Y = 0.9569·X + 33.24, R² = 0.37, p < 0.01) and Spearman correlation (ρ = 0.56, p < 0.01). The relationship between emicizumab plasma concentration and aPTT was well described by an inhibitory Emax model.

In the infant population, sustained FVIII-like activity and normalization of aPTT suggested a concentration-effect relationship, with higher emicizumab concentrations in infants compared to older individuals contributing to excellent bleeding control.

Concentration-Efficacy Relationships

The systematic review by Donners et al. performed detailed exposure-response modeling using an Emax model relating trough plasma concentrations to annualized bleeding rates. The EC50 for treated bleeds was 1.47 µg/mL with an Emax of 0.96, while for treated joint bleeds the EC50 was 1.09 µg/mL with an Emax of 0.98. Critically, the control of bleeds did not improve above emicizumab concentrations of 30 µg/mL, establishing an effectiveness plateau. At 30 µg/mL, the expected ABR was 2.4 for treated bleeds and 1.1 for treated joint bleeds, while at 50 µg/mL these values were 1.9 and 1.0, respectively.

The comprehensive exposure-response analysis by Jonsson et al. using a generalized Poisson distribution and inhibitory Emax relationship found that emicizumab concentrations were inversely related to bleeding rates, with higher concentrations leading to reduced bleeding. The IC50 was estimated at 3.58 µg/mL, indicating the concentration required for half-maximal effect. Importantly, concentrations above 30 µg/mL provided clinically meaningful control of bleeding. At the average steady-state concentration across all regimens (53.5 µg/mL), the predicted mean annualized bleeding rate was 1.28, corresponding to a 94.0% reduction from baseline. Simulations demonstrated that the three dosing regimens (1.5 mg/kg QW, 3 mg/kg Q2W, 6 mg/kg Q4W) maintain concentrations near the plateau of effect.

Visual predictive checks from the exposure-response model showed adequate prediction of bleeding onset over time, supporting the model’s validity. The PK profile of once-weekly emicizumab provided sustained therapeutic plasma levels consistent with population PK models.

Factors Affecting Pharmacokinetics and Pharmacodynamics

Body Weight and Size Effects

Body weight was a significant covariate affecting emicizumab pharmacokinetics across multiple studies. Dosing regimens were universally based on body weight (mg/kg), and including body weight in population PK models significantly reduced interindividual variability in both clearance (CL/F) and volume of distribution (V/F). Body weight also affected apparent clearance and distribution volume in the exposure-response model. The Chinese pediatric study documented body size characteristics, with median weight of 14.5 kg (range 8.5-41.8 kg) and median BMI of 16.8 kg/m² (range 13.8-21.5 kg/m²).

Plasma concentrations of emicizumab increased in a dose-dependent manner, and steady-state trough levels increased in a dose-proportional manner, supporting body weight-based dosing strategies. However, the systematic review suggested that individualized monitoring of emicizumab concentrations may allow for more cost-effective dosing.

Age Effects

Age emerged as an important covariate affecting emicizumab pharmacokinetics and potentially pharmacodynamics. Bioavailability decreased with age, particularly after 65 years. Conversely, infants demonstrated higher emicizumab concentrations compared to older individuals with the same dosing regimen, with concentrations of 60-65 µg/mL in HAVEN 7 versus 46.7 ± 14.9 µg/mL in older HAVEN 1-4 participants and 46-48 µg/mL in HAVEN 2-3.

The Chinese pediatric study population had a median age at enrollment of 3.52 years (range 0.93-16.25 years) and a median age at emicizumab initiation of 1.40 years, with height (median 92.0 cm) also documented. The HAVEN 7 studies focused specifically on infants ≤12 months of age, demonstrating feasibility and efficacy in this youngest population. Studies in adults included participants from age 12 to 75 years.

Inhibitor Status and Prior Treatment History

FVIII inhibitor status did not significantly affect emicizumab PK variability or efficacy. Emicizumab was effective in patients with FVIII inhibitors as well as those without, with no significant difference in bleeding control between the two groups.

Prior treatment history varied across studies and was considered as a potential covariate. Several studies included previously untreated or minimally treated patients (PUPs/MTPs). The Chinese pediatric study reported that 15 patients were previously untreated or minimally treated, and 9 had a history of inhibitors. In the exposure-response analysis, prior prophylaxis with FVIII significantly reduced bleeding hazard by 44.7%, representing the most clinically significant covariate effect identified.

Other Covariates

Albumin levels affected pharmacokinetics, with low albumin levels associated with decreased exposure and albumin affecting apparent clearance. The Chinese pediatric study documented median albumin levels of 46.1 g/L (range 41.4-49.6 g/L).

Ethnic differences were observed, with African race mentioned as a factor affecting PK variability and ethnicity affecting distribution volume in population PK models. The Chinese pediatric study specifically noted that ethnic and developmental factors—such as body size, nutritional status, and earlier treatment initiation—may influence drug exposure and clinical response.

Neutralizing antidrug antibodies represented a primary source of interindividual PK variability, though their actual occurrence was rare. Age >30 years was also identified as a factor affecting PK variability.

Safety and Immunogenicity

Emicizumab demonstrated a favorable safety profile across all studies, with most adverse events being mild and not leading to treatment discontinuation.

Injection Site Reactions and General Tolerability

Injection site reactions were the most frequent treatment-related adverse event. In the Phase I/II extension, mild injection site reactions occurred in 38.9% of patients (7/18). In the infant studies, 16.4-16.7% of participants experienced emicizumab-related injection site reactions, all Grade 1. The HAVEN 4 study reported injection site reactions in 22% of patients.

Most participants in infant studies experienced at least one adverse event (92.6-100%), but importantly, no adverse events led to treatment withdrawal, modification, or interruption. The long-term Phase I/II extension reported no discontinuations due to safety concerns, and the treatment was generally well tolerated with limited toxicity.

Thrombotic Events

Notably, no thromboembolic events or thrombotic microangiopathies occurred in the infant studies or the long-term Phase I/II extension. The HAVEN 4 study also observed no thrombotic events. In the HAVEN 1 study, two participants experienced thrombotic microangiopathy and thromboembolism, but these were associated with the use of bypassing agents rather than emicizumab itself. The systematic review noted an increased thrombotic risk when emicizumab was used with high doses of activated prothrombin complex concentrate.

Immunogenicity

Emicizumab demonstrated minimal immunogenicity across studies. None of the participants in the HAVEN 7 studies tested positive for anti-drug antibodies (ADAs) at any timepoint. The early Phase I study reported no development of antibodies to emicizumab.

In the Phase I/II extension, 4 patients tested positive for ADAs, but these were nonneutralizing and did not affect PK/PD parameters. Similarly, the systematic review noted that neutralizing antidrug antibodies affected PK variability but their actual development was rare. The HAVEN 4 study observed no development of de-novo antidrug antibodies with neutralizing potential.

Regarding development of FVIII inhibitors in previously untreated patients, one PUP in HAVEN 7 was confirmed positive for FVIII inhibitors after three FVIII exposure days, and another PUP tested positive after 10 FVIII exposure days.

Laboratory Safety Parameters

Emicizumab did not significantly affect key laboratory safety parameters. Platelet count, fibrinogen, D-dimer, and prothrombin fragment 1.2 (PF1.2) levels were not significantly affected. In the infant study, aPTT was normalized by Week 3 and mean FIX and FX concentrations were unaffected by emicizumab.

Serious Adverse Events

Serious adverse events were infrequent and not attributed to emicizumab. In HAVEN 7, 29.1% of participants reported 30 serious adverse events, all considered serious due to hospitalization, but none were considered emicizumab-related. The Phase I/II extension reported that 4 patients experienced serious adverse events, but none were related to emicizumab treatment. One participant in HAVEN 7 experienced an anaphylactic reaction following food allergy, not considered emicizumab-related.

Study Methodologies

The studies employed diverse methodological approaches to assess emicizumab pharmacokinetics and pharmacodynamics, ranging from Phase I dose-escalation studies to large Phase III randomized trials and sophisticated modeling analyses.

Study Designs

The evidence base included open-label, non-randomized, dose-escalation Phase I studies, Phase III multicenter, open-label, randomized studies, Phase IIIb multi-center, open-label studies, and a two-stage Phase III study with run-in and expansion cohorts. The systematic review pooled data from 15 studies including 469 patients with hemophilia A and 140 volunteers. The exposure-response modeling study analyzed pooled Phase III trial data. One real-world study employed a retrospective single-center design.

Pharmacokinetic Assessment Methods

PK sampling strategies varied by study. The HAVEN 1 study collected blood samples before the first dose, weekly for the first month, every 2 weeks for the second month, every 4 weeks from the third to sixth month, every 8 weeks from the seventh to twelfth month, and every 12 weeks thereafter. HAVEN 7 measured plasma trough emicizumab concentrations at Week 5 and thereafter. The Chinese pediatric study collected peripheral blood samples during both loading and maintenance phases.

Bioanalytical methods for emicizumab concentration measurement included validated enzyme-linked immunosorbent assay (ELISA) in Phase I-III studies and a modified one-stage clotting assay (OSA) in Phase IV studies.

Pharmacodynamic Assessment Methods

FVIII-equivalent activity was measured using chromogenic assays. Assay precision and accuracy for emicizumab concentration measurement were validated in key studies.

Study Limitations

Several methodological limitations were noted. The Phase I/II extension was not randomized or controlled and had small patient numbers in each dose cohort. Sample sizes in some studies were based on clinical considerations rather than statistical power calculations. The systematic review cited limitations including a limited number of available studies, heterogeneous study populations, and lack of blinded, placebo-controlled studies.

Synthesis

The pharmacokinetic and pharmacodynamic characteristics of emicizumab demonstrate remarkable consistency across diverse populations and study designs, supporting its use as a prophylactic treatment for hemophilia A. Several key findings emerge from synthesis of the evidence.

Dose-Concentration-Response Relationships

The evidence establishes a clear hierarchy of dose-proportional pharmacokinetics leading to sustained therapeutic concentrations, which correlate with pharmacodynamic biomarkers and translate to clinically meaningful bleeding control. The convergence of findings from mechanistic studies, clinical trials, and sophisticated modeling analyses provides robust evidence for a threshold effect, with bleeding control plateauing above concentrations of 30 µg/mL. This plateau has important implications for clinical practice.

Age-Related Pharmacokinetic Differences

The consistently higher emicizumab concentrations observed in infants compared to older populations represent a genuine population-specific difference. Infants have lower total body weight but proportionally higher body surface area and different tissue composition compared to adults, which may reduce volume of distribution relative to body weight.

Clinical Implications of PK/PD Variability

The moderate interindividual variability in trough concentrations supports body weight-based dosing without routine therapeutic drug monitoring. However, specific subpopulations may benefit from concentration monitoring due to factors like low albumin levels or the development of neutralizing antibodies.

Reconciling Safety Findings

The favorable safety profile, particularly the absence of thromboembolic events in most studies, indicates that the observed thrombotic events are associated with drug-drug interactions rather than an inherent property of emicizumab.