Pharmacokinetics of IL-23p19 Inhibitors: A Systematic Review Across Immune-Mediated Inflammatory Diseases

Abstract

Background

Interleukin (IL)-23p19 inhibitors are increasingly used in immune-mediated inflammatory diseases (IMIDs), particularly inflammatory bowel disease (IBD). Almost half of patients show inadequate response, partly due to pharmacokinetic (PK) variability. However, a comprehensive overview of their PK parameters is lacking, which limits understanding of population variability. This systematic review assessed the PK of IL-23p19 inhibitors in healthy subjects, IBD and other IMIDs.

Methods

A systematic literature search was performed in PubMed/MEDLINE and Embase. Studies were screened by two independent researchers. Quality was assessed using the Cochrane risk of bias tool and the Clinical Pharmacokinetic Study Checklist.

Results

A total of 21 studies were included (risankizumab n = 9, guselkumab n = 6, tildrakizumab n = 3, mirikizumab n = 3), covering healthy subjects, psoriasis, psoriatic arthritis, Crohn’s disease (CD), ulcerative colitis (UC) and pustular/erythrodermic psoriasis. Only one study examined a distinct IBD group, specifically, mirikizumab in pediatric UC. For risankizumab, dose-normalized maximum concentrations and area under the concentration‑time curves over a dosing interval, along with central and peripheral volume of distribution and clearance were generally similar across CD and UC. Patients with CD and UC demonstrated increased risankizumab trough levels over consecutive dosing intervals, while an opposite trend was observed for other IMIDs. Covariate analysis revealed that only albumin in UC and body weight in CD were relevant covariates to risankizumab PK.

Conclusion

Limited PK data and lack of subphenotype-specific data on IBD highlight the need for future studies to better define PK parameters. Also, more research is needed to guide personalized dosing strategies for IL-23p19 inhibitor use in IBD, as well as in other IMIDs.

Graphical Abstract

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1007/s40262-026-01693-7.

Plain Language Summary

Doctors can prescribe biologics to patients diagnosed with autoimmune disease. Not all patients respond to biologic therapy in the same way. In part, this is caused by their bodies processing the biologics differently. Therefore, it is important to study how biologics are processed by the human body. A new class of biologics, called IL-23p19 inhibitors, have already shown promising treatment results for patients with psoriasis and those with inflammatory bowel disease (IBD). Unfortunately, differences in treatment responses among patients are observed, making it difficult for doctors to know whether the treatment would be successful for their patients. In order to provide doctors with more insights in this topic, this study summarized all the current knowledge on ways in which IL-23p19 inhibitors are processed by the bodies of patients with autoimmune disorders, with a specific focus on patients with IBD. Our findings show that only limited research has been done for patients with IBD. Even less attention has been given to specific IBD subtypes. For example, research is scarce with regard to patients with Crohn’s disease who experience frequent perianal disease activity and fistulas. We also highlight the importance of personalized treatment and show how our study can help guide it. Understanding how IL-23p19 inhibitors are processed by the body can help doctors choose the best dose and time for each patient. This leads to treatments that are better tailored to the needs of specific patients, resulting in better treatment outcomes and increased quality of life for patients living with autoimmune diseases.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1007/s40262-026-01693-7.

Key Points

Little attention has been paid to ways in which IL-23p19 inhibitors are processed differently among patients with IBD.

Elaborating on this, even less attention has been directed to patients with specific IBD subtypes.

Future studies should focus on ways in which IL-23p19 inhibitors are processed by patients with IBD. Subsequently, this knowledge should be used to guide personalized treatment approaches.

Introduction

Immune-mediated inflammatory disorders (IMIDs), such as inflammatory bowel disease (IBD), including Crohn’s disease (CD) and ulcerative colitis (UC), and psoriasis, are characterized by pro-inflammatory cytokine signaling involved in disease pathogenesis [1, 2]. Cytokines such as tumor necrosis factor (TNF)-α, interleukin (IL)-6, IL-12 and IL-23 have become important therapeutic targets for monoclonal antibodies, commonly referred to as biologics [3]. Among biologics, IL-23p19 inhibitors have recently shown promising efficacy and safety profiles across Phase 3 clinical trials in psoriasis, active psoriatic arthritis (PsA), CD, and UC [4–14].

Interleukin-23 is a pro-inflammatory cytokine of the IL-12 family consisting of two subunits, p19 and p40 [15]. Activation of the IL-23/Th17 axis promotes the production of downstream pro-inflammatory cytokines, such as IL-17 and IL-22 [16, 17]. Selective inhibition of the p19 subunit blocks IL-23–mediated inflammatory processes, while preserving IL-23–independent pathways [15–17]. Since IL-23 plays a central role in the pathogenesis of IBD, psoriasis and active PsA, these IMIDs represent the main indications for IL-23p19–targeted treatment [18].

Consequently, several IL-23p19 inhibitors have been approved by the Food and Drug Administration (FDA) and European Medicines Agency (EMA). Guselkumab gained approval from both the FDA and EMA in 2017, followed by tildrakizumab in 2018, risankizumab in 2019 and mirikizumab in 2023 [19–26]. Both risankizumab and guselkumab are approved for CD, UC, psoriasis, and active PsA, whereas mirikizumab is limited to CD and UC and tildrakizumab is approved for psoriasis only [19–26].

All these agents are monoclonal antibodies, which influence their mechanisms of action and administration routes. Due to their low stability and limited absorption in the gastrointestinal tract, IL-23p19 inhibitors have to be administered either intravenously (IV) or subcutaneously (SC) [27]. Moreover, biologics have high interindividual pharmacokinetic (PK) variability [27]. Variation in drug responses often arises because of patient-specific factors, such as the presence of antidrug antibodies (ADAs) and body weight [28, 29]. This variability in PK is associated with poorly predictable drug efficacy [30]. A better understanding of PK and the factors influencing it, may improve treatment outcomes and ultimately enhance patients' quality of life [31].

This systematic review aims to summarize available literature on relevant PK parameters and dosing regimens of risankizumab, tildrakizumab, guselkumab and mirikizumab in healthy individuals and in patients with IMIDs, including IBD. Although the primary interest of this systematic review relies on IBD, other IMIDs are also included, especially since PK data from these diseases might provide valuable insights due to overlap in pathophysiology and patients who suffer from more than one IMID. This, in combination with exposure-response relationships, provides a foundation for the development of personalized treatment approaches, including therapeutic drug monitoring (TDM), aiming to optimize the use of IL-23p19 inhibitors.

Methods

Search Strategy

This systematic review was conducted according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines [32]. The study was registered in the international prospective register of systematic reviews (PROSPERO) with protocol number CRD420251155343. The systematic literature search was performed using PubMed/MEDLINE and Embase and took place on June 16, 2025. Only articles written in English were included, regardless of publication date. Search terms included 'pharmacokinetics', ‘risankizumab’, ‘tildrakizumab’, ‘guselkumab’ and 'mirikizumab'. Details of search terms for both PubMed/MEDLINE and Embase can be found in Supplementary materials ‘Search terms’.

Selection Criteria

All search results were imported into Rayyan [33]. Rayyan was used to detect duplicates, which were manually removed by one researcher (IAP). Titles and abstracts were screened to identify potentially relevant studies focusing on risankizumab, tildrakizumab, guselkumab and mirikizumab or other IL-23p19 inhibitors in clinical trials, as well as the presence of PK data and exposure. Next, a full-text assessment was performed to determine whether the study met the inclusion criteria. Inclusion was based on: (i) individuals who were treated with risankizumab, tildrakizumab, guselkumab or mirikizumab and (ii) studies which focused on PK parameters (absorption, metabolism, elimination, and distribution) or exposure. Reasons to exclude studies were: (i) animal studies; (ii) article types such as reviews, case reports and surveys; (iii) no full-text availability and (iv) concomitant use of multiple biologics in the same patients. Both the title and abstract screening, as well as the full-text assessment, were done independently by two researchers (IP and AG). All potential discrepancies were flagged by Rayyan and were resolved through discussions and examined by a third researcher (PM) when consensus could not be reached.

Data Extraction

Data extraction from all included studies was performed by a single investigator (IP). All data were independently checked by a second (AG) and third investigator (PM). The extracted study characteristics included: study design, type and characteristics of the biologic and population details. Extracted PK data of exposure encompassed absorption, concentrations, distribution, metabolism and excretion, type of PK model, covariates and dose recommendations. Data visualization was performed using GraphPad 10.6.1 and all maximum concentration (C_max), minimum concentrations (C_min) and area under the curve (AUC) values were dose normalized. The graphical abstract was created using BioRender.

Quality Assessment

The Cochrane risk of bias tool (RoB 2 tool) and the Clinical Pharmacokinetic Study Checklist (ClinPK checklist) were used to assess the quality of evidence [33, 34]. Both tools have been independently assessed by two researchers (IAP and AG). Studies were not excluded based on their risk of bias.

Results

Study Selection and Data Extraction

The study selection procedure is shown in the PRISMA flow diagram (Fig. 1). This systematic review included 21 studies reporting the PK or population (pop)PK of IL-23p19 inhibitors [35–55]. Study designs and cohorts are detailed in Table S1 and dosing schemes in Fig. 2. Key popPK parameters extracted are summarized in Tables S2–S3. All PK models for risankizumab, tildrakizumab, and guselkumab used non-compartmental analysis [35, 39, 42, 44, 45, 47, 49]. The following sections describe PK parameters followed by popPK parameters, for each IL-23p19 inhibitor.

Fig. 1.

PRISMA flow diagram illustrating the study selection process for studies assessing the PK parameters of IL-23p19 inhibitors. CYP cytochrome P450, IL interleukin, PK pharmacokinetics, PRISMA preferred reporting items for systematic reviews and meta-analyses

Fig. 2.

Schematic visualization of different dosing schemes for risankizumab, tildrakizumab, guselkumab, and mirikizumab as were reported in the included studies. CD Crohn’s disease, EP erythrodermic psoriasis, GPP generalized pustular psoriasis, IV intravenous, PBO placebo, PPP palmoplantar pustulosis, PsA psoriatic arthritis, q every, SC subcutaneous, UC ulcerative colitis, w weeks

Quality of Evidence

In total, six randomized controlled trials (RCTs) were included, which were all rated as high quality using the Rob 2 tool (Fig. S1). The remaining 15 studies were assessed using the ClinPK checklist. Of these, three were judged to have a high risk of bias, ten with an intermediate risk and two with a low risk of bias (Fig. S2).

Risankizumab

Nine of the included studies reported the PK of risankizumab [35–43]. Psoriasis was addressed in 10 studies, healthy subjects and patients with CD in six studies, active PsA in four studies and UC, generalized pustular psoriasis (GPP) and erythrodermic psoriasis (EP) in one study each. The number of identified PK studies may differ from the number of studies within a specific IMID, as PK studies or popPK models typically use pooled data from multiple studies. In addition to Fig. 2, the effects of several dosing schemes for healthy subjects were evaluated across White, Japanese and Chinese populations. Moreover, two studies in healthy subjects compared syringe pumps (SP), prefilled syringe pumps (PFS), on-body injectors (OBI) and autoinjector administrations [39, 42]. All popPK models were two-compartmental [36–38, 40, 41, 43]. Most studies did not report whether the data exhibited apparent target-mediated disposition (TMDD). One study indicated that no apparent TMDD was observed in healthy subjects and patients with psoriasis, GPP or EP [38].

Absorption

Considering the assessed PK parameters, the time to reach maximum plasma concentration (T_max) varied from 5 to 7 days, irrespective of device type (range 2–42) [39, 42]. Assessment of the extent of absorption demonstrated an SC bioavailability of 59% [35]. The relative bioavailability of the 150 milligram (mg)/milliliter (mL) PFS was comparable to the bioavailability of the 90-mg/mL as well as the 150-mg/mL autoinjector (AI) (Table S2) [39]. A comparable trend was observed for PFS versus OBI [42].

From popPK models, derived rate of absorption (K_a) ranged from 0.121 (relative standard error [RSE] 0.896%) day⁻¹ in CD to 0.230 (RSE 4.8%) day⁻¹ in healthy subjects and patients with psoriasis, GPP or EP [38, 41]. Absolute bioavailability was consistent across Phase 3 studies, with reported values of 0.89 (RSE 7.2%) [37, 38]. In healthy subjects, T_max did not vary across ethnicities or SC dosing [38].

Concentrations

The C_max/dose values were generally consistent across different dosing regimens and administration methods, with a notably higher C_max/dose observed for 90 mg/mL PFS administered as two SC injections (Fig. S3A).

Across popPK analysis, steady state was reached at week 16 for healthy subjects and patients with psoriasis, GPP and EP [37]. Moreover, despite observed differences in C_max/dose across healthy White, Chinese and Japanese subjects, the overall trends remain inconclusive (Fig. 3A). The C_max/dose showed a modest increase after successive doses for patients with CD or UC (Fig. 3A) [41, 43]. Dose normalized minimum concentrations (C_min) (trough levels) increased over successive dosing during induction therapy in patients with CD and UC, while they decreased for patients with psoriasis, active PsA, GPP and EP (Fig. 3B).

Fig. 3.

Model derived PK parameters of risankizumab in healthy subjects and patients with psoriasis, active PsA, GPP, EP, CD or UC following IV and SC doses. (A) C_max/dose values estimated by NCA in patients with psoriasis and under different dosing techniques in healthy subjects. (B) Observed trough levels in healthy White, Chinese, and Japanese subjects and in patients with active PsA, CD or UC. (C) V_d values estimated by NCA in healthy subjects and patients with psoriasis, active PsA, GPP, EP, CD or UC. (D) AUC/dose, AUC_∞/dose and AUC_last/dose values estimated by NCA in healthy subjects using different dosing techniques. (E) AUC_∞/dose and AUC_τ /dose values derived from popPK models in healthy White, Japanese and Chinese subjects and in patients with active PsA, CD or UC. AI autoinjector, AUC area under the plasma drug concentration–time curve, AUC_∞ area under the plasma concentration–time curve from time zero to infinity, AUC_last area under the curve up to the last measurable concentration, AUC_τ area under the plasma concentration–time curve over a dosing interval, CD Crohn’s disease, C_max maximum concentration, C_min minimum concentration observed (trough level), EP erythrodermic psoriasis, GPP generalized pustular psoriasis, IV intravenous, K_a absorption rate constant, L liter, Min minutes, NCA noncompartmental analysis, OBI on-body injector, PFS prefilled syringe, PK pharmacokinetics, pop population, PsA psoriatic arthritis, S seconds, SC subcutaneous, SP syringe pump, UC ulcerative colitis, V_d volume of distribution, V_c central volume of distribution, V_p peripheral volume of distribution, W week

Distribution

Regarding PK parameters, the reported terminal phase volume of distribution (V_d) for patients with psoriasis was 10.8 L [35]. Dose normalized area under the curve (AUC) values are shown in Figure 3D. The 90 mg/mL PFS administered as two SC injections showed higher AUC_∞/dose and AUC_last/dose compared to 150 mg/mL AI and 150 mg/mL PFS [39]. Similarly, 90 mg/mL administered as four SC injections showed higher AUC_last/dose and AUC_∞/dose [42]. The 216 mg/2.4 mL SP injected within ≤ 6 min showed slightly lower AUC_last/dose and AUC_∞/dose [42].

In the popPK analysis, the central volume of distribution (V_c) was generally similar across diseases (Fig. 3C). However, in a study that included patients with both psoriasis and CD, V_c was observed to be slightly higher. The peripheral volume of distribution (V_p) was generally lower in patients with CD and UC compared to psoriasis, active PsA, GPP and EP (Fig. 3C). The AUC_∞/dose and AUC_τ/dose are shown in Fig. 3E. The AUC_∞/dose decreased with increased SC doses for White, Chinese, and Japanese subjects [38]. The AUC_∞/dose was comparable for 600 mg IV and 1200 mg IV, but higher than 200 mg IV in Japanese subjects. The AUC_τ/dose increased after successive doses for CD and UC [41, 43].

Metabolism and Excretion

Across PK parameters, clearance (CL) of 0.33 L/day was reported [35]. Apparent clearance (CL/F) values were generally similar across different PFS and syringe pump dosing regimens, with an elevated value for 216 mg/2.4 mL syringe pump injected within six minutes (Fig. S3B). For all dosing schemes and techniques, half-life (t_1/2) varied between harmonic mean 21.7 (pseudo-stable diffusion [SD] 8.59) days and harmonic mean 30.9 (pseudo-SD 7.03) days [39, 42]

In context of popPK parameters, CL was higher in studies that included healthy subjects or patients with CD and in studies that included patients with psoriasis or CD compared to studies that included healthy subjects or patients with psoriasis, active PsA, GPP or EP (Figure S3C). Substantial differences in intercompartmental clearance (Q) were observed with the lowest values in patients with CD (0.255 [RSE 4.20%] L/day) and patients with UC (0.254 [RSE 5.40%] L/day) and the highest in active PsA (0.839 [RSE 8.2%] L/day) [40, 41, 43]. The t_1/2 after SC administration did not differ between healthy White, Japanese and Chinese subjects [38]. Following IV dosing in Japanese subjects, a similar t_1/2 of harmonic mean (HM) 30.7 (pseudo-SD 2.56) to HM 32.7 (pseudo-SD 14.2) days was observed [38]. Studies with healthy subjects and patients with active PsA or psoriasis showed a t_1/2 of 26.3 days and 28 days, respectively [37, 38].

Covariates Associated with PK Variability

Two PK studies showed that ADAs do not interfere with risankizumab exposure [39, 42]. One study found that ADA (n = 2, p = NR) had a minimal effect on risankizumab PK [35]. In popPK analysis, body weight was found to positively correlate with CL (0.93 [RSE 10%]), V_C and V_P (0.99 [RSE 11%]) in patients with both psoriasis and CD, but near the lower bound [36]. The ADA titers ≥ 128 caused a proportional increase in CL of 43% [37, 38]. Patients with psoriasis treated with 150-mg SC risankizumab and with body weight > 100 kg were predicted to have 30% lower exposures, C_max and AUC compared to patients with body weight < 100 kg [37]. In patients with UC, only the negative association between albumin and CL was clinically meaningful and in patients with GPP or EP only body weight correlated with risankizumab [38, 43]. None of the tested covariates had a (clinically relevant) association with risankizumab in patients with active PsA [40].

Dosing Options

Several dosing options were provided, but in the absence of predefined therapeutic targets. Risankizumab 150 mg/mL PFS, 90 mg/mL PFS, and 150 mg/mL AI showed comparable bioequivalence in PK studies and could therefore be considered as potential administration doses and methods [39, 42]. The PopPK simulations were used to analyze what exposures can be expected under various dosing regimens, thereby informing for possible dosing regimens. Patients with psoriasis or active PsA could be treated with 150 mg SC at Weeks 0 and 4, and every 12 weeks thereafter [37]. For patients with GPP or EP, no dose adjustments based on ethnicity would be needed [38]. For patients with CD, 600 mg risankizumab IV at Weeks 0, 4, and 8 is suggested for induction of remission followed by maintenance treatment of 360 mg SC every eight weeks [41]. Lastly, for patients with UC, induction of remission could be achieved by 1200 mg IV every four weeks, followed by maintenance dosing of 180 mg or 360 mg SC every eight weeks starting at week 12 [43].

Tildrakizumab

Three studies were included reporting the PK of tildrakizumab [44–46]. Two studies included only healthy subjects, whereas one study included both healthy subjects and patients with psoriasis. In addition to Fig. 2, one study compared different dosing regimens among healthy White, Japanese and Chinese subjects [44]. Two of the studies focusing on tildrakizumab reported PK parameters, whereas one study reported popPK parameters [44, 45]. The popPK study used a one-compartmental model [46].

Absorption

Across the PK parameters, T_max varied depending on the route of administration. In case of IV administration, T_max was approximately 0.08 days, while in the case of SC administration a T_max ranging from 6.03 to 13.00 days was observed [44]. A bioavailability of 92% was reported for a single SC dose of tildrakizumab, a little higher than previously reported [25, 44].

Across the popPK parameters, K_a was 0.458 (RSE 6.8%) day⁻¹. The absorption lag time was almost similar between healthy subjects (1.19 [RSE 7.7%] hours) and patients with psoriasis (1.2 hours) [46]. Bioavailability was 31% higher in healthy subjects compared to patients with psoriasis [46].

Concentrations

The C_max/dose showed similar values among White and Japanese subjects after increased SC doses (Fig. 4A). Moreover, reported C_max/dose values were generally consistent for different IV doses, with a lower value observed for Japanese subjects.

Fig. 4.

Model derived PK parameters of tildrakizumab in healthy subjects and healthy White and Japanese subjects following IV and SC doses. (A) C_max/dose values estimated by NCA in healthy subjects and healthy White and Japanese subjects. (B) AUC_∞/dose and AUC_last /dose values estimated by NCA in healthy subjects and healthy White and Japanese subjects. (C) V_d, V_d/F and V_z/F values estimated by NCA in healthy White, Chinese and Japanese subjects. AUC area under the plasma drug concentration–time curve, AUC_∞ area under the plasma concentration–time curve from time zero to infinity, AUC_last area under the curve up to the last measurable concentration, C_max maximum plasma concentration, IV intravenous, NCA noncompartmental analysis, PK pharmacokinetics, pop population, SC subcutaneous, V_d volume of distribution, V_d/F apparent volume of distribution after extravascular administration, V_z/F apparent volume of distribution during the terminal phase after extravascular administration

In a popPK analysis, steady state (ss) was reached by week 16 under the clinical dosing regimen [46]. A C_max,ss was achieved in 8.1 (34%) days for 100 mg SC and in 16.3 (33%) days for 200 mg SC.

Distribution

The V_d was similar across IV doses in healthy subjects (Fig. 4B). Additionally, no notable differences were observed in the apparent volume of distribution (V_d/F) after extravascular administration for 50 mg SC and 200 mg IV in healthy subjects. Moreover, the V_d/F during the terminal phase (V_z/F) was consistent across White, Chinese and Japanese subjects. AUC_last/dose and AUC_∞/dose were consistent within each route of administration but differed between routes (Fig. 4C).

Across popPK parameters, V_d did not differ across studies (Fig. S4A). The AUC_ss increased linearly with increasing dosing (100 mg SC: 305 [CV 41%] µg*day/mL, 200 mg SC: 612 [CV 40%] µg*day/mL) [46].

Metabolism and Excretion

The CL/F increased with higher SC doses in healthy subjects (50 mg: 197 mL/day [CV 27%]; 200 mg: 282 mL/day (CV 119%)) and was similar across White, Chinese and Japanese subjects (158–165 mL/day; %CV 13.7–22.4) [44, 45]. CL for IV dosing ranged from 2.04 (CV 23%) mL/day/kg to 2.52 (CV 15%) mL/day/kg depending on the administered dose [45]. The t_1/2 for IV administration varied between 24.6 (CV 25%) days and 29.7 (CV 20%) days [45]. In case of SC administration, t_1/2 was approximately 25 days irrespective of dosing (50 mg SC: 25.1 [CV 12%] days, 200 mg SC: 24.5 [CV 22%] days) [45].

In popPK analysis, CL was similar across studies for patients with psoriasis (Figure S4B) [46]. Following SC administration, t_½ absorption was 1.5 (CV 18%) days, whereas t_½ elimination was 23.4 (CV 23%) days [46].

Covariates Associated with PK Variability

Among the different studies, several covariates were integrated in the final model Table S2 and S3. However, none of these covariates were clinically relevant.

Dosing Options

Several dosing options are proposed, without predefined therapeutic targets. SC administration was tested in White and Chinese subjects and IV administration was tested in Japanese subjects, with similar PK outcomes [44]. The PopPK analysis indicated that no dose adjustments would be required based on covariates [46].

Guselkumab

A total of six studies assessed the PK parameters of guselkumab [47–52]. Nine studies reported on patients with psoriasis, five on patients with active PsA, two studies on patients with PPP and two studies on healthy subjects. Again, the number of identified PK studies may differ from the number of studies within a specific IMID, as PK studies or popPK models typically use pooled data from multiple studies. In addition to Fig. 2, one study looked at the use of different guselkumab formulation types in healthy subjects including 100 mg IV or 100 mg SC in a lyophilized formulation, a pre-filled syringe uncoated (PFS-U) or a pre-filled syringe with an integrated device (PFS-FID) [51]. One popPK study reported a two-compartment model while three reported a one-compartment model structure [48, 50–52].

Absorption

The T_max varied for a single SC administration of 10, 30, 100 or 300 mg in patients with psoriasis [47, 49]. The lowest median T_max was 3.2 days (range 2.0–7.0) at 100 mg SC, while the highest was 6 days (range 4.0–13.2) at 10 mg SC. In a study including patients with both psoriasis and active PsA, a median T_max of 4.0 (range 2.97–13.99) days for 10 mg SC was observed [49]. The T_max was similar across 30 mg, 100 mg and 300 mg SC, with medians of approximately 6 days (ranges 2.97–13.99) [47].

Across popPK analyses, reported K_a differed considerably with values ranging from 4.93 (RSE 4.9%) day⁻¹ to 1.11 (RSE 14.1%) day⁻¹ for patients with psoriasis and to 0.572 (RSE 8.69%) day⁻¹ for patients with active PsA [48, 50, 52]. A study which included healthy subjects and patients with psoriasis, active PsA or PPP, reported a K_a of 0.313 day⁻¹ [51]. This study also described an absolute bioavailability of 49.2%.

Concentrations

The C_max/dose for healthy subjects and patients with both psoriasis and active PsA were generally similar regardless of dose and administration route (Fig. 5A). Variation in C_min (trough levels) was observed (Fig. 5B).

Fig. 5.

Model-predicted PK parameters for guselkumab in healthy subjects and patients with psoriasis, active PsA and PPP following IV and SC doses. (A) C_max/dose values estimated by NCA in healthy subjects and patients with psoriasis or active PsA. (B) Observed trough levels in healthy subjects and patients with psoriasis, active PsA and PPP. (C) V_d and V_d/F values estimated by NCA in healthy subjects and patients with psoriasis. (D) AUC_∞/dose and AUC_last/dose values estimated by NCA in healthy subjects and patients with psoriasis and active PsA. (E) AUC_ss/dose values derived from popPK models in healthy subjects and patients with psoriasis, active PsA and PPP. AUC area under the plasma drug concentration–time curve, AUC_∞ area under the plasma concentration–time curve from time zero to infinity, AUC_last area under the curve up to the last measurable concentration, AUC_ss area under the plasma drug concentration–time curve at steady state, CL clearance, C_max maximum plasma concentration, C_min minimum concentration observed (trough level), IV intravenous, NCA noncompartmental analysis, PK pharmacokinetics, pop population, PPP palmoplantar pustulosis, PsA psoriatic arthritis, SC subcutaneous, V_d volume of distribution, V_d/F apparent volume of distribution

Distribution

The V_d was generally similar across different IV doses in healthy subjects and patients with psoriasis (Fig. 5C). The V_d/F was slightly higher in patients with psoriasis receiving 10 mg SC or 300 mg SC compared to patients receiving 30 mg or 100 mg SC. The AUC_last/dose and AUC_∞/dose were generally consistent among different doses and administration routes except for healthy subjects receiving IV guselkumab (Fig. 5D). Here, the AUC_∞/dose decreased with increasing IV doses.

Across popPK analysis, V_d/F was consistent across patients with psoriasis and patients with active PsA (Fig. S5A). The central volume of distribution (V_c) was higher than the peripheral volume of distribution (V_p) in a study with healthy subjects and patients with psoriasis, active PsA or PPP. Figure 5E showed that AUC_{week 0/8, SS} values were highly comparable between study groups.

Metabolism and Excretion

Following IV administration in healthy subjects, CL was slightly lower at higher doses (1–10 mg/kg; 3.62–4.61 mL/day/kg) compared with lower doses (0.03–0.3 mg/kg; 5.52–6.03 mL/day/kg) (Fig. S5B). The CL/F was higher for 10 mg and 300 mg SC compared to 30 mg and 100 mg SC. Intravenous administration of guselkumab in healthy subjects resulted in t_1/2 varying from 12.3 ± 1 days (0.03 mg/kg) to 1 mg/kg: 19.1 ± 2.2 days (1 mg/kg) [47]. Subcutaneous administration of 10–300 mg SC in a study including patients with psoriasis and patients with active PsA, resulted in similar t_1/2 between 15.6 ± 3.0 and 17.6 ± 3.1 days [49].

In the popPK analyses, comparable CL/F values were reported for patients with psoriasis and patients with active PsA (Fig. S5C). A t_1/2 of 18.1 days was reported for patients with psoriasis [50].

Covariates Associated with PK Variability

No covariates were reported in the PK analyses. Across popPK analysis, body weight was found to correlate with guselkumab PK across two studies involving patients with psoriasis [48, 50]. Both CL/F and V/F were found to scale approximately linearly with body weight with reported exponents of 1.16 (RSE 11.1%) and 0.998 (RSE 4.37%) for CL/F and 1.04 (RSE 13.1%) and 0.829 (RSE 4.61%) for V/F.

Dosing Options

No potential dose recommendations were provided from the PK analyses. Based on the popPK analyses, several potential dosing schemes were considered as possible treatment doses. However, no clear predefined therapeutic targets were defined. One study suggested that no dose adjustments based on disease type (psoriasis, active PsA or PPP) are needed [51]. Another study considered 100 mg SC guselkumab at Weeks 0 and 4 and every 8 weeks thereafter for patients with active PsA as a possible treatment regimen [52]. No dose adjustments based on diabetes status or body weight would be necessary for patients with active PsA [52].

Mirikizumab

Three studies reported on the popPK of mirikizumab [53–55]. Healthy subjects were addressed in one study, whereas UC was addressed in two studies. All included studies reported popPK analysis, whereas none reported standard PK analyses. Two studies used two-compartment models [53, 55]. One study used a non-compartmental analysis [54].

Absorption

Comparable K_a values were reported for patients with UC in the AMAC trial and in the LUCENT 1 and 2 studies, with mean values of 0.18168 (95% confidence interval [CI] 0.16368–0.22152; standard error of the estimate [SEE] 5.63%) days⁻¹ and 0.16944 (95% CI 0.12888–0.23952; SEE 17.3%) days⁻¹, respectively [53]. Healthy subjects reported an absolute bioavailability of 42.8% for 200 mg SC mirikizumab and 34.2% for 400 mg SC mirikizumab [54]. In the AMAC trial, patients with UC reported a bioavailability of 42.2% (95% CI 36.8–49.1; SEE 4.5%) [53]. With regard to the LUCENT 1 and 2 studies, a bioavailability of 47.6% (95% CI 34.5–59.5; SEE 15.9%) was reported [53]. A SC bioavailability of 49.8% (95% CI 40.6–59.1; SEE 9.48%) was observed for pediatric patients with UC [55]. In healthy individuals, T_max was 2.98 days and did not vary across different SC doses of 200 mg (median 2.98 [range 2.97 −6.95] days) and 400 mg (median 2.98 [range 2.97 −7.04] days) [54].

Concentrations

In healthy subjects, C_max/dose values were consistent within each administration route, but differed between the IV and SC administration routes (Fig. 6A). Pediatric patients with UC with weights > 40 kg showed a C_max/dose comparable with adults for IV and SC administration.

Fig. 6.

Model-predicted popPK parameters of mirikizumab in healthy subjects and adult and pediatric patients with UC following IV and SC doses. (A) C_max/dose values derived from popPK models in healthy subjects and patients with UC. (B) V_c, V_p and V_z values derived from popPK models observed in healthy subjects and adult and pediatric patients with UC. (C) AUC_∞/dose and AUC_last/dose values derived from popPK models in healthy subjects. (D) CL values derived from popPK models observed in healthy subjects and adult and pediatric patients with UC. AUC area under the plasma drug concentration–time curve, AUC_∞ area under the plasma concentration–time curve from time zero to infinity, AUC_last area under the plasma concentration–time curve from time zero to the time of the last measurable concentration, CL clearance, C_max maximum plasma concentration, IV intravenous, NCA noncompartmental analysis, PK pharmacokinetics, pop population, SC subcutaneous, UC ulcerative colitis, V_c central volume of distribution, V_d volume of distribution, V_p peripheral volume of distribution, V_z volume of distribution during the terminal phase

Distribution

The V_c and V_p were quite similar between adult patients with UC and pediatric patients with UC (Fig. 6B). The volume of distribution during the terminal phase (V_z) increased for increasing IV dose in healthy subjects and V_ss was similar for both 600 mg and 1200 mg IV doses, but slightly higher compared to 300 mg IV. The AUC_last/dose and AUC_∞/dose were largely comparable for IV doses, with only a slight decrease observed for 1200 mg IV (Fig. 6C). Similarly, AUC_last/dose and AUC_∞/dose were generally consistent for SC dose, with a modest decrease observed for the 400-mg SC dose.

Metabolism and Excretion

Patients with UC in the AMAC trial showed a CL of 0.552 (95% CI 0.5232–0.6312; SEE 2.94%) L/day, which was comparable with the LUCENT 1 and 2 studies, where a CL of 0.528 (95% CI 0.4608–0.6; SEE 6.96%) L/day was reported [53]. Patients with UC in the AMAC trial showed similar CL values compared to patients with UC in LUCENT 1 + 2 studies (Fig. 6D). Moreover, pediatric patients showed a comparable CL to that observed for adult patients with UC. The CL varied for different IV and SC doses in healthy subjects. Moreover, inter-compartmental clearance (Q) yielded similar findings among the AMAC trial (0.18144 L/day [95% CI 0.13536–0.4896; SEE 17.6%]) and the LUCENT 1 and 2 studies (0.2088 L/day [95% CI 0.08448–0.4512; SEE 42.2%]) [53].

Covariates Associated with PK Variability

Although albumin was found to be a covariate influencing mirikizumab PK, it was not clinically relevant in adult and pediatric patients with UC [53–55]. Moreover, none of the other tested covariates were found to be clinically relevant associated with the PK of healthy subjects or adult patients with UC [54, 55].

Dosing Options

No dose adjustments based on body weight and body mass index (BMI) are warranted for adult patients with UC [53]. For pediatric UC patients, dosing could be generally based on body weight. During induction therapy, 300 mg IV is considered for patients weighing > 40 kg or 5 mg/kg IV every four weeks for patients weighing ≤ 40 kg [55]. Following maintenance therapy, potential doses could be 200 mg SC every four weeks for patients weighing > 40 kg, 100 mg SC for those weighing > 20–≤ 40 kg and 50 mg SC for those weighing ≤ 20 kg [55]. Again, these dosing options were provided without predefined therapeutic targets.

Discussion

This systematic review is the first to investigate the currently available literature on PK parameters of IL-23p19 inhibitors in adults and children across IMIDs. By doing so, we hope to contribute to a deeper understanding of treatment variability, driven by PK. Our review highlights that only a few Phase 1 to 3 studies and no Phase 4 studies focused on the PK parameters of these biologics, rendering their use in advanced treatment approaches such as model-informed precision dosing (MIPD). Based on this evidence, we draw several conclusions for clinicians and pharmacometricians, as outlined below.

Key PK Parameters Show Marked Variation Across IMIDs

The PK of IL-23p19 inhibitors showed marked variations across IMIDs. First, patients with both psoriasis and active PsA who received risankizumab showed decreased trough levels over successive dosing intervals [37, 38]. Interestingly, an opposite trend was observed for patients with CD and UC [40, 41, 43]. These observations may suggest that disease-specific pathophysiology (e.g., inflammatory burden, protein loss or receptor expression) may contribute to PK variability. In particular, for IBD, altered intestinal barrier integrity, fecal drug loss and systemic inflammation are known to associate with clearance, suggesting that exposure-response relationships established in dermatological indications may not directly translate [56]. Previous studies have consistently shown that body weight interferes with the clearance of monoclonal antibodies [29]. Higher body weight and BMI were observed in patients with both psoriasis and active PsA compared to patients with CD and UC. However, covariate analysis showed that body weight associated with risankizumab exposures, C_max and AUC only in patients with psoriasis [37]. Moreover, in both studies focusing on patients with psoriasis, ADA titers ≥ 128 were associated with increased risankizumab clearance, which may have potentially led to lower trough levels over successive dosing intervals [37, 38]. This finding of ADAs interfering with PK of monoclonal antibodies has been widely reported in the literature [29]. Our systematic review indicates that these previous described interactions also apply to risankizumab.

Second, V_p was generally lower in patients with CD and UC compared to patients with psoriasis who received risankizumab. Due to their large molecular structure and limited ability to diffuse into tissues, biologics tend to stay in the central compartment, resulting in a generally low V_p for biologics [29]. Although V_p was within the normal ranges for both diseases, the observed difference could be explained by tissue penetration [57]. One explanation could be that patients with psoriasis generally have relatively higher skin permeability compared with the gut of patients with CD and UC, resulting in a slightly higher V_p.

Third, the bioavailability of risankizumab was 31% higher in healthy subjects compared to patients with psoriasis [46]. Previous work suggests that this is largely attributed to the covariate healthy volunteers and, to a lesser extent, to body weight [58].

Fourth, large differences in K_a were found between patients with psoriasis and patients with active PsA who received guselkumab. Consequently, high intra-individual variability (IIV) magnitudes were reported, reflecting the complexity of SC absorption as well as the interaction with numerous covariates influencing K_a. For example, previous studies have shown that the injection site, injection depth and skin morphology can influence the K_a [57]. All of these characteristics vary substantially among patients. Limited studies addressed patient-specific factors interfering with the K_a of IL-23p19 inhibitors. However, understanding these differences is essential for optimizing patient-specific treatment approaches. Therefore, further research focusing on factors affecting the K_a of IL-23p19 inhibitors is required to enable individualized dosing strategies and techniques.

Fifth, a relatively high IIV on bioavailability in pediatric patients with UC who received mirikizumab was observed [55]. Since weight differences were accounted for through allometric scaling, the observed IIV must be attributable to other factors. A review article on the PK of biologics in children argues that extracellular fluid and lymphatic flow are increased in younger children compared to older children and adults, thereby contributing to variability in bioavailability [59]. However, only one group received mirikizumab SC, with a mean age of 14 years, reducing the potential influence of extracellular volume and lymphatic flow. In addition to age-related factors, IIV of bioavailability may also associate with concomitant use of immunomodulators or corticosteroids, which was not reported in the reviewed articles. These agents can modulate immunogenicity and thereby alter monoclonal antibody exposure [60]. It would therefore be relevant to study the use of concomitant immunomodulators or corticosteroids as covariates in popPK models for IL-23p19 inhibitors.

Optimizing Dosing Regimens of IL-23p19 Inhibitors in Specific Diseases and Populations

Only a few disease-specific PK studies of IL-23p19 inhibitors have been performed. Within the context of IBD, it would be relevant to assess PK behavior under conditions of acute severe inflammation, such as acute severe ulcerative colitis (ASUC) and in disease phenotypes such as perianal CD, where altered drug distribution or loss may occur. For example, patients with ASUC have demonstrated PK profiles that complicate TDM for infliximab treatment, since their PK make TDM more complex [61]. Additionally, a study on ustekinumab in patients with perianal disease suggested that these patients could benefit from TDM [62]. Integration of real-world PK data with TDM may accelerate the refinement of dosing strategies in these special populations.

Given the sparsity of disease-specific PK studies, very few studies have focused on special populations. For example, this systematic review included 21 studies, of which only one specifically studied a pediatric population. Recently, the FDA approved guselkumab for the treatment of pediatric plaque psoriasis and active PsA, marking an important milestone in biologic therapy in pediatric patients [63]. The PROTOSTAR study has already demonstrated that guselkumab is both effective and well tolerated in pediatric patients, making it interesting to also study the use of other IL-23p19 inhibitors in pediatric patients [64]. Despite emerging pediatric approvals, harmonized PK bridging studies across age groups are lacking. In contrast to the pediatric population, studies focusing on PK of pregnant women remain notably understudied. One systematic review on the PK of biologics throughout pregnancy already demonstrated that the anatomical and physiological changes throughout pregnancy may have effects on the PK of biologics, underscoring the need for further research on IL-23p19 inhibitors throughout pregnancy [65].

Covariates Influencing PK Variability

Although four of the seven PK studies and 13 of 14 popPK studies reported covariates that could potentially impact the PK of IL-23p19 inhibitors, only a few demonstrated clinically meaningful effects [37, 43, 48, 50]. This represents part of a larger knowledge gap in the literature. Currently, no standardized criteria exist for assessing the clinical significance of covariate effects on the PK of monoclonal antibodies. Consequently, the choices are largely based on physiology, clinical judgment and covariates collected during the study [66]. A previous study emphasizes the importance of establishing a clear definition of clinically relevant covariates, due to their ability to explain variability in drug exposure and response [67]. Furthermore, beyond classical covariates, more dynamic biomarkers (e.g., albumin, inflammatory proteins and fecal calprotectin) could serve as predictors of clearance in future models. Future studies should aim to develop standardized guidelines for evaluating the clinical relevance of covariates and establish thresholds to realize personalized dosing approaches [68]. As mentioned before, several covariates were identified to have a clinically relevant effect on the PK of risankizumab and guselkumab. These covariates should be validated in independent cohorts and should then be considered for integration into MIPD. Although not clinically relevant, clearance was found to be associated with risankizumab. No definitive explanation for this correlation has been established.

Dosing Options

Several dosing regimens can be considered for treatment with IL-23p19 inhibitors, once they have been evaluated in exposure-response studies. In case of risankizumab, reported doses for patients with psoriasis, active PsA and UC were consistent with currently approved induction and maintenance dosing regimens [69, 70]. In patients with CD, a maintenance dose of 180 mg SC was not reported in our systematic review, but this is approved as maintenance therapy [69, 70]. Current potential doses for tildrakizumab were solely based on healthy subjects. Although data on tildrakizumab dosing based on PK parameters have already been described for patients with psoriasis, these have not yet been published [58]. Reported doses of guselkumab for patients with active PsA were in line with current guidelines [71, 72]. Additionally, the EMA reports that patients at high risk of joint damage, based on clinical judgment, may benefit from 100 mg administered every four weeks [71]. However, this was not identified in our systematic review. Furthermore, no specific dosing recommendations based directly on PK parameters for adult patients with CD or UC were identified. We did find a proposed dosing recommendation for pediatric patients with UC using mirikizumab [55]. Here, induction dosing was divided into > 40 kg (300 mg IV) and ≤ 40 kg (5 mg/kg IV). Maintenance dosing was proposed in three groups: ≤ 20 kg (50 mg SC), > 20 to ≤ 40 kg (100 mg SC) and > 40 kg (200 mg SC). Nevertheless, mirikizumab has not yet been approved for use in pediatric populations and additional trials are warranted to confirm the efficacy and safety of mirikizumab. Finally, the suggested doses raise several concerns. First, most studies that suggested potential dosing regimens originated from industry-sponsored trials, often without validation in independent datasets. Second, the reported dosing recommendations often lack a direct foundation on therapeutic targets. Therefore, future studies are urgently needed that focus on defining therapeutic targets for IL-23p19 inhibitors, which can then be used in future PK studies. For instance, such predefined therapeutic targets can be incorporated in future PK-guided trials and will help to evaluate whether dose intensification may benefit subgroups with accelerated clearance.

Strengths and Limitations

Our systematic review has several key strengths. First, Phase 1, 2 and 3 studies were included, resulting in a comprehensive overview of popPK across different stages of clinical development. This systematic synthesis of popPK parameters across agents may provide a basis for future cross-disease comparative modelling. Second, the included RCTs were of high methodological quality, as demonstrated by a low risk of bias according to the Rob 2 tool. Third, we revealed substantial knowledge gaps in PK data and key covariates of IL-23p19 inhibitors in adult and pediatric patients with psoriasis, active PsA, GPP, PPP, CD, and UC. Several limitations also need to be addressed. Since data of clinical trials were used across multiple studies, there was a general lack of original study data contributing to popPK analysis. Second, three studies were ranked as a high risk of bias and ten studies as an intermediate risk of bias. A probable explanation could be that these studies were not primarily designed to study PK parameters, which makes them less suitable for evaluation with the Clin PK Checklist. Third, although covariates were tested, some studies combined data from patients with different indications and healthy subjects, leading to limited generalizability in a specific disease group. Fourth, very limited data were available across gastrointestinal indications relative to dermatological indications, precluding disease-specific recommendations for IBD. Further research should focus on better defining PK parameters of IL-23p19 inhibitors in specific disease and populational contexts and integrating them in popPK models to enable personalized dosing strategies.

Clinical Implications

From a clinical perspective, understanding PK variability in IL-23p19 inhibitors is particularly relevant as these agents are predominantly used in patients with prior biologic failure, where altered PK may contribute to primary or secondary non-response. This has been demonstrated for TNF-antagonists [30]. Identifying patient-specific factors such as body weight, inflammatory burden, albumin and ADAs may therefore help to individualize dosing schedules. Incorporating PK/PD data into treatment algorithms could enhance therapeutic precision, reduce under-exposure and also improve durability of response. Moreover, as mentioned before, it is important to realize that such models should be tailored to specific diseases and preferably to specific subtypes, since they cannot be easily extrapolated to other patient populations. By doing so, more effective treatment approaches can be developed for pediatric and pregnant patients. Moreover, incorporating disease-specific indications into MIPD models enables their use for both gastrointestinal and dermatological indications.

Conclusion

This systematic review on PK parameters of IL-23p19 inhibitors in healthy subjects and patients with IMIDs aimed to provide a comprehensive overview of currently available PK data. Across included studies, the PK of IL-23p19 inhibitors is predominantly studied in healthy subjects, with a concomitant lack of disease-specific PK studies and data across special populations. Additionally, only a limited number of clinically meaningful covariates have been reported, hindering the development of MIPD for IL-23p19 inhibitors. Bridging PK insights with exposure-response and real-world outcome data represents the next critical step toward MIPD in IBD and other IMIDs. Finally, potential dosing regimens were reported despite the lack of predefined therapeutic targets. Further research is therefore warranted to better define the PK parameters of IL-23p19 inhibitors, to enable personalized dosing strategies and to better define the position of IL-23p19 inhibitors in the therapeutic landscape of its primary indications.

Supplementary Information

Below is the link to the electronic supplementary material.

Funding

No sources of financial assistance were used to assist the preparation of the manuscript or to conduct the study. ARB is supported by the EU Horizon Health grant ID-DarkMatter-NCD (101136582) and the Innovation Health Initiative Joint Undertaking (IHI-JU) grant INTERCEPT (101194780). GD is supported by the EU Horizon Europe Program grant miGut-Health: personalized blueprint of intestinal health (101095470). JT is supported by Innovation Health Initiative Joint Undertaking (IHI-JU) grant INTERCEPT (101194780) and the GlycanTrigger Grant (101093997). Views and opinions expressed are however those of the author(s) only and do not necessarily reflect those of the European Union or the European Research Council. Neither the European Union nor the granting authority can be held responsible for them. The funders had no role in study design, data collection and analysis, preparation of the manuscript or decision to publish.

Declarations

Conflict of interest

The corresponding author confirms on behalf of all authors that there have been no involvements that might raise the question of bias in the work reported or in the conclusions, implications, or opinions stated. RJK is shareholder of Insight RX, San Francisco, CA, USA. MCD reports receiving consultancy and advisory fees from AbbVie, Abivax, AstraZeneca, Boehringer Ingelheim International GmbH, Bristol-Meyer Squibb, Celltrion, Eli Lilly and Company, F. Hoffmann-La Roche Ltd, Janssen Pharmaceuticals, Johnson and Johnson, Merck, Pfizer Inc, Prometheus Labs, Sanofi, Spyre, and Takeda Pharmaceuticals, and licensing fees from Takeda Pharmaceuticals. JT has received consulting/advisory board fees from Abbvie, Pfizer, Janssen, Tillots Pharma, Takeda, Sandoz and Lilly; grant support from Janssen and Abbvie. RA has served as a speaker, consultant, or received research grants from AbbVie, Abivax, AlfaSigma, AstraZeneca, Bristol-Myers Squibb, Celltrion Healthcare, Galapagos, Johnson & Johnson, Lilly, MSD Sharp & Dohme, Pfizer, and Takeda Pharma. PMa is a full-time employee of Ionis and may hold Ionis stock or options. ED received consultancy fees from Alimentiv and Argenx; lecture fees from Celltrion and Galapagos; and financial support from Celltrion, Janssen, Pfizer, Prometheus Biosciences, R-Biopharm, and Sandoz, outside the submitted work, with all honoraria/fees being paid to KU Leuven and not to any personal account of ED. GD received a grant from Royal DSM, advisory board fees from Pharmacosmos and Astra-Zeneca, and speakers fee from Abbvie. HJV received grants paid to the University Medical Center Groningen from Albireo/Ipsen and Mirum and consultancy fees paid to the University Medical Center Groningen from Albireo/Ipsen, Intercept, Mirum, Orphalan, ProQR and Vertex, all outside of the submitted work. ARB reports receiving research grants from Janssen Pharmaceuticals and received speaker’s and advisory board fees from AbbVie and Ferring Pharmaceuticals, outside the submitted work. All other authors have no conflicts of interest to declare.

Availability of data and material

Not applicable. All available data are included within the manuscript and supplementary files.

Ethics approval

Not applicable.

Consent to participate

Not applicable.

Consent for publication

Not applicable.

Code availability

Not applicable.

Author contributions

Conceptualization: PM and ARB. Data curation: IAP, AG, PM and ARB. Formal analysis: IAP, AG and PM. Funding acquisition: PM and ARB. Investigation: all authors. Methodology: IAP, AG, PM and ARB. Project administration: IAP and ARB. Data acquisition: IAP, AG and PM. Resources: IAP, AG and PM. Sources: NA. Supervision: PM and ARB. Validation: all authors. Visualization: IAP and AG. Writing—original draft: IAP, AG, PM and ARB. Writing—review and editing: all authors. All authors read and approved the final version.