Pharmacokinetics of Certolizumab Pegol in Pregnancy: Results from the Open-Label CHERISH Study

Abstract

Introduction

Control of chronic inflammatory diseases (CIDs) during pregnancy is essential for maternal and foetal health; however, the impact of pregnancy on the pharmacokinetics (PK) of CID therapies is unknown. This study investigated the impact of pregnancy on the PK of certolizumab pegol (CZP), a tumour necrosis factor inhibitor (TNFi), in women with CIDs.

Methods

CHERISH (NCT04163016) was a multicentre, longitudinal, open-label phase 1B study evaluating the impact of pregnancy on the PK of CZP in women with CIDs. Pregnant participants on a stable CZP regimen were enrolled at ≤ 10 weeks of gestation. The primary variable was pre-dose and post-dose plasma CZP concentrations throughout pregnancy and postpartum in patients who received ≥ 1 dose. Treatment-emergent adverse events (TEAEs) were recorded.

Results

Of 21 enrolled participants (CZP 200 mg every 2 weeks [Q2W], n = 15; CZP 400 mg Q2W, n = 1; CZP 400 mg Q4W, n = 5), 16 completed the study. Relative to postpartum, pre-dose plasma CZP concentrations were modestly reduced across trimesters 1–3, with no clear pattern in post-dose concentrations. TEAEs occurred in 81.0% of participants, with ‘infections and infestations’ being most common; only 1 (4.8%) was considered treatment related. Five participants (23.8%) experienced serious TEAEs; none were considered treatment-related. One serious TEAE of foetal death in a high-risk twin pregnancy, one spontaneous abortion was reported in an enrolled participant before their first dose of CZP in the study, and no infant illnesses were reported.

Conclusions

CZP plasma concentrations were modestly lower during pregnancy versus postpartum, were consistent across trimesters, and were within the range observed in studies of non-pregnant individuals with CIDs. Safety was consistent with the established profile of CZP and the patient population. Findings support maintenance of CZP dosing regimens during pregnancy.

Trial Registration

ClinicialTrials.gov identifier: NCT04163016.

Supplementary Information

The online version contains supplementary material available at 10.1007/s40744-026-00853-2.

Key Summary Points

Why carry out this study?

Control of chronic inflammatory diseases (CIDs) during pregnancy is essential for maternal and foetal health; however, the impact of pregnancy on the pharmacokinetics (PK) of CID therapies is poorly understood.

This open-label, phase 1B study sought to understand the impact of pregnancy on the pharmacokinetics of certolizumab pegol (CZP) in women with CIDs.

What was learned from the study?

Observed CZP plasma concentrations were lower during pregnancy versus postpartum but consistent across trimesters, and safety was consistent with the established profile of CZP.

Findings support the maintenance of established CZP dosing regimens during pregnancy, in order to maintain stable plasma concentrations.

Introduction

Most chronic inflammatory diseases (CIDs) are more common in women than men, and disease onset frequently overlaps with peak reproductive age [1, 2]. Accordingly, approximately 50% of women with autoimmune diseases, such as rheumatoid arthritis (RA), psoriatic arthritis (PsA), plaque psoriasis (PsO), Crohn’s disease and axial spondyloarthritis (axSpA), require therapeutic intervention during pregnancy, taking into account the benefits of disease control and the risks of various therapies [2]. For women living with these CIDs, maternal disease activity is correlated with adverse pregnancy outcomes, therefore physicians and patients may decide to continue treatment for optimal disease control during pregnancy [3–6]. However, clinical studies typically exclude pregnant women and withdraw patients who become pregnant, creating a lack of data on drug exposure in pregnancy and leaving treating physicians with limited evidence regarding therapeutic dosing during pregnancy [7]. Hence, there is an unmet need to establish the impact of pregnancy on the pharmacokinetics (PK) of CID therapeutics such as tumour necrosis factor inhibitors (TNFis).

Previous studies on the use of therapeutic agents for various CIDs have demonstrated varying effects of pregnancy on their pharmacokinetics (PK) [8, 9]. For example, it has been shown that the PK of hydroxychloroquine and azathioprine for rheumatic disease can vary during the course of pregnancy, potentially affecting therapeutic response in pregnant individuals [10–13]. A study of patients with inflammatory bowel disease reported that, when adjusting for the individuals’ changes in serum albumin, body mass index (BMI) and C-reactive protein (CRP), the maternal serum levels of the TNFi infliximab increased during pregnancy, while adalimumab levels remained consistent [14].

Certolizumab pegol (CZP), a recombinant, humanised, monoclonal antigen-binding antibody fragment (Fab′) that is conjugated to polyethylene glycol, is a fragment crystallisable region (Fc)-free TNFi, approved in adults for the treatment of CIDs, including Crohn’s disease, RA, PsA, PsO and axSpA, in > 60 countries worldwide [15–17]. CZP is administered subcutaneously at 200 mg every 2 weeks (Q2W) or 400 mg every 4 weeks (Q4W) and has been demonstrated to exhibit ~ 80% absolute bioavailability following subcutaneous administration, with a terminal elimination half-life of approximately 14 days [15]. Importantly, CZP is the only TNFi with clinical trial data demonstrating minimal-to-no placental transfer, suggesting CZP may be a suitable choice of TNFi during pregnancy [2, 18]. The lack of placental transfer is due to the absence of an IgG Fc region in CZP, which differs from other TNFis, and prevents binding to the neonatal Fc receptor (FcRn), thus minimising active FcRn-mediated transfer across the placenta [19, 20].

At present, however, for women who continue CZP therapy during pregnancy, there is a paucity of data regarding the impact of pregnancy on CZP PK. There is, therefore, a need for further data on whether pregnancy-related physiologic changes impact the PK of CZP. Here, we present findings from the CHERISH study, which sought to investigate the longitudinal PK and safety of the use of CZP during pregnancy and postpartum by measuring maternal plasma CZP concentration and monitoring treatment-emergent adverse events (TEAEs).

Methods

Study Design and Participants

The CHERISH study (NCT04163016; study protocol available at ClinicalTrials.gov) was a multicentre, longitudinal, prospective, open-label phase 1B exploratory study (conducted from July 2020–May 2023 in the United States, France, Germany, the Netherlands, Spain and Switzerland) evaluating the impact of pregnancy on the PK of CZP in women with CIDs. The study consisted of a screening period, pregnancy period (up to 40 weeks), postpartum period (up to 13 weeks), and a safety follow-up (5 weeks after final study visit) (Fig. 1). Only women of childbearing potential already receiving commercial CZP could register their interest in the study. After confirmation of pregnancy, women who consented to participate were eligible to enroll once they had been on a stable, maintenance dose of CZP for at least 12 weeks, at which point it was expected that CZP levels had reached steady state. The study was interventional due to the collection of blood samples from the participants beyond routine clinical practice. However, all other study assessments were part of routine clinical practice (except from the dosing diary; see Study Procedures), including regular concomitant medication and procedures reviews, adverse event (AE) review, weight and pregnancy outcome. Participant race and ethnicity were self-reported by patients through selection from a fixed set of categories.

Fig. 1.

Study design and sampling schedule. ^aWoCBA patients already on CZP could register interest in the study. After confirmation of pregnancy and consent to participate, women were eligible to enroll in the study once they had stable maintenance dosing for at least 12 weeks, at which point it was expected that CZP levels had reached steady state. ^bPregnancy period: PK samples were collected prior to the subsequent dose (pre-dose) Q4W starting with the first dose after enrolment. Further, post-dose (7 ± 1 day after dose administration) PK samples were collected Q8W throughout the pregnancy. It was expected that there would be approximately six pre-dose and three post-dose PK samples collected per participant (if there were no discontinuations) over the pregnancy. Postpartum period: A pre-dose sample 12 weeks (± 1 week) postpartum was collected during the postpartum period. If possible, one post-dose sample 1 week (± 1 day) after the 12 weeks postpartum dose was also collected. ^cSafety follow-up: All participants were contacted via telephone 5 weeks (± 5 days) after the final study visit. If any participant withdrew early, they completed the safety follow-up. ^dTiming and the number of visits across the pregnancy varied for individual participants. CZP certolizumab pegol, PK pharmacokinetics, Q2W every 2 weeks, Q4W every 4 weeks, Q8W every 8 weeks, WoCBA women of childbearing age

Eligible participants were ≥ 18 years of age at the time of enrolment, were pregnant (≤ 10 weeks of gestation at time of enrolment) and on a stable, commercially approved, maintenance dose CZP treatment for ≥ 12 weeks prior to being enrolled in the study. Additionally, participants required a negative interferon gamma release assay or tuberculin skin test within the prior 6 months, and there had been no change in the study participant’s clinical status, or social, family, or travel history within the prior 6 months. Key exclusion criteria included any medical condition or history (including obstetrical or psychiatric conditions) that, in the opinion of the investigator, could jeopardise or would compromise the pregnancy or the participant’s ability to participate in the study. Exclusions were also applied for women who received medication in the prior 12 weeks with strong positive evidence of human foetal risk of teratogenicity during pregnancy and excluded women with a recent history of alcohol or drug abuse. Participants with evidence of clinically significant anaemia, significant documented morbidity in the past 12 weeks, significant drug allergies or severe post-treatment hypersensitivity reactions were also excluded.

Participants were free to withdraw from the study at any time. The decision to continue or discontinue CZP treatment during pregnancy was made between the participant and their treating physician, taking into account the potential risks to the mother, the benefits of CZP therapy, and according to the applicable local label.

The CHERISH study protocol, amendments, and patient informed consent were reviewed by a national, regional, or Independent Ethics Committee (IEC) or Institutional Review Board (IRB) (Table S4). CHERISH was conducted in accordance with the current version of the applicable regulatory and International Conference on Harmonisation (ICH)-Good Clinical Practice (GCP) requirements, the ethical principles that have their origin in the principles of the Declaration of Helsinki, and the local laws of the countries involved. Full details of the ethics committee involved are available upon request. All patients provided written informed consent to participate in the study, with the addition of obtaining parental consent where relevant.

Study Procedures

Participants were responsible for obtaining and administering commercially available CZP dosing regimens (200 mg Q2W or 400 mg Q2W or 400 mg Q4W) following the guidance of their treating physician. Participants were provided with a diary to record all CZP dosing dates and times over the study. The dosing regimen could be modified by the participant’s treating physician after entry into the study if the regimen did not exceed 400 mg Q2W, and the dose was recorded in the diary.

During the pregnancy period, pre-dose PK samples were collected Q4W, ≤ 4 h prior to dosing, starting with the first dose after enrolment. Post-dose PK samples were taken every 8 weeks (Q8W) throughout pregnancy, 7 (± 1) days after CZP dose administration, as this was anticipated to coincide with peak plasma concentrations. During the postpartum period, one pre-dose sample (12 [± 1] weeks postpartum) and one optional post-dose sample 7 (± 1) days after the 12 weeks postpartum dose were also collected. All participants were contacted via telephone for safety follow-up 5 weeks (± 5 days) after the final study visit. Participants who withdrew early still completed the safety follow-up (Fig. 1).

Plasma CZP concentrations were determined by a validated bioanalytical method using electrochemiluminescence (ECLIA) on a Meso Scale Discovery (MSD)® platform, with a lower limit of quantification (LLOQ) for CZP in plasma of 0.032 mg/l. Values below LLOQ were replaced by the value of LLOQ/2 (0.016 μg/ml) in the calculations of geometric mean and confidence intervals (CIs). A total of 239 plasma samples were analysed for CZP concentration across 12 batches, with an inter-run coefficient of variation for low, medium and high concentration quality control samples of 8.0–12.2%. The linear range of this assay was 32–5000 ng/ml. Anti-drug antibody (ADAb) titre was also measured with MSD ECLIA assay.

Outcomes

The primary objective of this study was to assess systemic CZP exposure across the course of pregnancy in patients with CIDs. This was assessed using pre-dose and post-dose plasma CZP concentrations in patients receiving ≥ 1 dose of CZP after enrolment (pharmacokinetic per-protocol set [PK-PPS]). The plasma CZP concentrations throughout pregnancy were compared with postpartum plasma CZP concentrations to allow comparison with the non-pregnant state. Secondary outcomes were the assessment of the formation of ADAbs throughout the study, TEAEs from time of informed consent through safety follow-up, and pregnancy outcome in the safety set, which included all enrolled participants who were already receiving commercial CZP at time of screening (and had been on CZP for at least 12 weeks prior to enrolment), as per the inclusion criteria.

Prespecified TEAEs included: serious infections, malignancies, congestive heart failure, demyelinating-like disorders, aplastic anaemia, pancytopenia, thrombocytopenia, neutropenia, and leukopenia, serious bleeding events, lupus and lupus-like syndrome, and serious skin reactions. Abnormal pregnancy outcomes (such as spontaneous abortion, foetal death, stillbirth, congenital anomalies and ectopic pregnancy) were considered serious adverse events (SAEs).

Statistical Analysis

Since no formal hypothesis testing was conducted, a sample size of 15 participants was considered appropriate for a phase 1B clinical pharmacology study and for the exploratory analyses performed; 15 evaluable participants were expected to result in a minimum of ten participants providing longitudinal data. Longitudinal data was defined as ≥ 4 quantifiable pre-dose and ≥ 2 quantifiable post-dose PK samples during pregnancy and ≥ 1 quantifiable postpartum PK sample. It was expected that up to 25–30 pregnant women with CIDs would need to be screened to ensure that approximately 20 participants were enrolled, resulting in approximately 15 evaluable participants.

The primary outcome of systemic CZP exposure across the course of pregnancy was assessed by estimating the adjusted least-squares mean differences and 95% CI for overall plasma CZP concentrations during pregnancy vs postpartum. This employed a linear mixed effects model (PROC MIXED) with disease phenotype as a time-invariant fixed effect, and gestational week, trimester and whether the sample was pre-dose or post-dose as time-varying effects; the random effect was the participant.

Due to the model assumptions and small sample size, a robust sandwich variance estimator was not used. The average pre-dose (or post-dose) results within each trimester per participant were used to calculate the geometric mean and 95% CI for all participants combined. Geometric mean was only calculated if at least 2/3 of the concentrations were quantified at the respective timepoint and n ≥ 4. The summary statistics were based on participant means within trimesters, where trimester 1 was defined as up to 12 weeks and 6 days of gestation, trimester 2 was defined as 13–28 weeks and 6 days of gestation, and trimester 3 was defined as any time at or after 29 weeks of gestation. Due to low patient numbers in the CZP 400 mg Q2W and Q4W groups, observed plasma CZP concentration data are only presented for the CZP 200 mg Q2W dose group; however, the overall model included all data at all dosing regimens (including both pre-dose and post-dose data). CZP plasma concentrations were additionally analysed adjusting for CRP, albumin and BMI as time-varying exploratory outcomes using the same model as the primary analysis, but with CRP, albumin and BMI as covariates in three separate models.

The ADAb sample status was determined for each visit where samples were collected. The number and proportion of dosed participants who were ADAb positive over time are reported.

All AEs were coded using the Medical Dictionary for Regulatory Activities (MedDRA) v26.0 by System Organ Class (SOC) and Preferred Terms (PT) and characterised as treatment emergent; TEAEs were summarised by intensity and by relationship to CZP. Statistical analysis and generation of tables, figures, participant data listings, and statistical output were performed using statistical analysis system (SAS)^® Versions 9.3–9.4.

Results

Patient Disposition and Baseline Characteristics

Twenty-six participants were screened and 21 participants started the study, (CZP 200 mg Q2W, n = 15; CZP 400 mg Q2W, n = 1; CZP 400 mg Q4W, n = 5), of which 16 (76.2%) completed the study. Of the 5 participants who discontinued the study, the reasons for study discontinuation were AE (n = 1), withdrawal by participants (n = 2) and other reasons (n = 2; n = 1 each for indeterminate tuberculosis test and participant feeling overwhelmed). The one participant in the CZP 400 mg Q2W group had their dose reduced to CZP 200 mg Q2W. This participant was only counted once under the initial dose group.

The mean age of participants was 32.2 (range: 24–40) years and the majority of participants were white (n = 20); the remaining participant was Asian. The most commonly reported primary indication for CZP was RA (n = 9 [42.9%]), followed by PsA (n = 4 [19.0%]), axSpA (n = 3 [14.3%]), Crohn’s disease (n = 2 [9.5%]), ankylosing spondylitis (n = 2 [9.5%]), and PsO (n = 1 [4.8%]; Table 1). The use of concomitant medications was reported in 20 participants (95.2%), most commonly folic acid (n = 8 [38.1%]), followed by cholecalciferol and paracetamol (n = 7 [33.3%] each), minerals, multivitamins and acetylsalicylic acid (n = 6 [28.6%] each) and omeprazole, plain vitamins, ferrous sulphate, ibuprofen, and sulfasalazine (n = 3 [14.3%] each).

Table 1.

Baseline demographics and disease characteristics (enrolled set)

Category	Dose group			All dosed participantsb N = 21
	CZP 200 mg Q2W n = 15	CZP 400 mg Q2Wa n = 1	CZP 400 mg Q4Wb n = 5
Age, years at screening
Mean (SD)	32.2 (4.7)	NAc	31.2 (4.2)	32.2 (4.6)
Median (min, max)	31.0 (24, 40)	NAc	31.0 (25, 36)	31.0 (24, 40)
Race, n (%)
Asian	1 (6.7)	0	0	1 (4.8)
White	14 (93.3)	1 (100)	5 (100)	20 (95.2)
Country, n (%)
France	2 (13.3)	0	0	2 (9.5)
Germany	5 (33.3)	0	0	5 (23.8)
Netherlands	4 (26.7)	0	0	4 (19.0)
Spain	0	0	1 (20.0)	1 (4.8)
Switzerland	1 (6.7)	1 (100)	2 (40.0)	4 (19.0)
United States	3 (20.0)	0	2 (40.0)	5 (23.8)
BMI,d kg/m2
Mean (SD)	28.0 (5.5)	NAc	25.0 (4.8)	27.1 (5.3)
Primary indication,e n (%)
Rheumatoid arthritis	7 (46.7)	1 (100)	1 (20.0)	9 (42.9)
Psoriatic arthropathy	3 (20.0)	0	1 (20.0)	4 (19.0)
Axial spondyloarthritis	0 2 (13.3)	0	1 (20.0)	3 (14.3)
Crohn’s disease		0	2 (40.0)	2 (9.5)
Ankylosing spondylitis	2 (13.3)	0	0	2 (9.5)
Psoriasis	1 (6.7)	0	0	1 (4.8)
Any past/prior/concomitant medicationf	14 (93.3)	1 (100)	5 (100)	20 (95.2)

^aOne participant had their dose reduced from CZP 400 mg Q2W to CZP 200 mg Q2W. This participant was only counted once under the initial dose group

^bOne participant was not dosed while enrolled in the study; this participant subsequently re-enrolled and was included in the CZP 400 mg Q4W group and the All dosed participants group

^cNot applicable replaced summary statistics when the sample size satisfied the n = 3 or n < 3 rule

^dBMI was measured at enrolment when pregnancy stage varied between participants and may have thus influenced measurements

^eThis summary included medical conditions that were identified as the “Primary Condition” for CZP treatment on the case report form

^fPast medications are those that are ongoing at the time of the first day of the last menstrual period and those that started after the first day of the last menstrual period that stopped prior to entering the screening period; prior medications are those that are ongoing at the time of the first day of the last menstrual period and those that started after the first day of the last menstrual period prior to entering the screening period; concomitant medications are medications with at least 1 day in common with the screening or sampling periods

BMI body mass index, CZP certolizumab pegol, kg kilogram, max maximum, mg milligram, min minimum, NA not applicable, Q2W every 2 weeks, Q4W every 4 weeks, SD standard deviation

Plasma CZP Concentrations and Immunogenicity

In the CZP 200 mg Q2W group (n = 15), the observed geometric mean of plasma CZP concentrations across all trimesters ranged from 15.3–16.9 μg/ml pre-dose and 20.0–25.0 μg/ml post-dose (Fig. 2). Postpartum, mean pre-dose and post-dose CZP concentrations were 22.3 μg/ml and 30.2 μg/ml, respectively.

Fig. 2.

Observed geometric mean of plasma CZP concentration in participants in the CZP 200 mg Q2W group by pregnancy trimester and postpartum (PK-PPS). The pre-dose and post-dose plasma CZP concentrations are not presented for the CZP 400 mg Q2W and Q4W groups as there were insufficient individual data. Geometric mean was only calculated if at least 2/3 of the concentrations were quantified at the respective timepoint and n ≥ 4. The summary statistics were based on participant means within trimesters, where trimester 1 was defined as up to 12 weeks and 6 days of gestation, trimester 2 was defined as 13–28 weeks and 6 days of gestation, and trimester 3 was defined as any time at or after 29 weeks of gestation. Values below limit of quantification were replaced by the value of LLOQ/2 = 0.016 μg/ml in the calculations of geometric mean and CIs. CI confidence interval, CZP certolizumab pegol, LLOQ lower limit of quantification, mg milligram, PK-PPS Pharmacokinetic Per-Protocol Set, Q2W every 2 weeks, Q4W every 4 weeks, µg microgram, ml millilitres

Based on the observed data, in the CZP 400 mg Q4W group (n = 5), the geometric mean of plasma CZP concentrations in trimesters 2 and 3 were 9.0 μg/ml and 2.8 μg/ml pre-dose, and 46.1 μg/ml and 8.1 μg/ml post-dose (the low geometric mean for the value is due to the small sample size [n = 4] comprising one value below the lower limit of quantitation), respectively. There were insufficient data to calculate the geometric mean of plasma CZP concentrations during trimester 1 and postpartum. For the CZP 400 mg Q2W group, the geometric mean of pre-dose and post-dose plasma CZP concentrations at any trimester and postpartum were not calculated as there were insufficient individual data (n = 1 in this group).

Based on the primary analysis linear mixed effects statistical modelling, when including both pre-dose and post-dose plasma CZP concentrations, plasma CZP concentrations were lower during pregnancy (trimesters 1 through 3) relative to postpartum. When the linear mixed effects model was fitted to the pre-dose and post dose CZP concentrations separately, pre-dose plasma CZP concentrations were found to be lower during all 3 trimesters of pregnancy compared with postpartum, however, there was no clear pattern in post-dose plasma CZP concentrations (Fig. 3).

Fig. 3.

Mean plasma CZP concentration changes (µg/ml) in all dosed participants relative to postpartum by trimester from linear statistical mixed effects model (PK-PPS). Mean changes from postpartum plasma CZP concentration were derived from primary analysis linear mixed model contrasts (overall model included both pre-dose and post-dose data). CI confidence interval, CZP certolizumab pegol, PK-PPS Pharmacokinetic Per-Protocol Set, µg microgram, ml millilitres

The linear mixed effects model was adjusted for albumin, CRP, and BMI, as these variables reflect physiological changes during pregnancy that could be related to the observed lower CZP concentrations in pregnancy compared to postpartum. Adjustment for albumin and BMI resulted in a reduction in the least squares mean difference in CZP concentrations for pregnancy vs postpartum, but there was no impact of adjusting for CRP (Fig. 4).

Fig. 4.

Mean plasma CZP concentration changes (µg/ml) in all dosed participants relative to postpartum adjusted for serum albumin, CRP and BMI by trimester from mixed model (PK-PPS). Mean changes from postpartum plasma CZP concentration were derived from primary analysis linear mixed model contrasts. BMI body mass index, CI confidence interval, CRP C-reactive protein, CZP certolizumab pegol, PK-PPS Pharmacokinetic Per-Protocol Set, µg microgram, ml millilitres

The incidence of ADAb positivity among dosed participants in the study was 85.7% at enrolment, generally remained high throughout pregnancy, and was 75.0% postpartum (Supplementary Material: Supplementary Table 1). There was no pattern of change in ADAb titre classification over the course of pregnancy.

Safety

Of all enrolled participants, 17 (81.0%) experienced a TEAE (Table 2). Most of the TEAEs reported were in the System Organ Class of ‘Infections and Infestations’ (n = 12 [57.1%]). The most common TEAEs by Preferred Term included COVID-19 (n = 6 [28.6%]) and headache (n = 3 [14.3%]). Other TEAEs included mastitis, influenza, nasopharyngitis, cystitis, RA (worsening), migraine, postpartum haemorrhage, and gestational diabetes (n = 2 [9.5%] each) (Supplementary Table 2). All other TEAEs were reported in less than two participants. The most common TEAEs by trimester and postpartum are presented in Supplementary Table 3; there was no clear pattern of TEAEs across pregnancy trimesters.

Table 2.

Incidence of TEAEs by dose group (SS)

Category	Dose group			All dosed participantsd N = 21 n (%) [#]a,b
	CZP 200 mg Q2W n = 15 n (%) [#]a,b	CZP 400 mg Q2Wc n = 1 n (%) [#]a,b	CZP 400 mg Q4W n = 5 n (%) [#]a,b
Any TEAEs	13 (86.7) [57]	1 (100) [1]	3 (60.0) [8]	17 (81.0) [66]
Serious TEAEse	4 (26.7) [4]	0	1 (20.0) [1]	5 (23.8) [5]
Metabolism and nutrition disorders	1 (6.7) [1]	0	0	1 (4.8) [1]
Hyponatremia	1 (6.7) [1]	0	0	1 (4.8) [1]
Pregnancy, puerperium, and perinatal conditions	2 (13.3) [2]	0	1 (20.0) [1]	3 (14.3) [3]
Gestational diabetes	0	0	1 (20.0) [1]	1 (4.8) [1]
Vascular disorders	1 (6.7) [1]	0	0	1 (4.8) [1]
Haemorrhage	1 (6.7) [1]	0	0	1 (4.8) [1]
Severe TEAEsf	2 (13.3) [2]	0	0	2 (9.5) [2]
Premature rupture of membranes	1 (6.7) [1]	0	0	1 (4.8) [1]
Foetal death	1 (6.7) [1]	0	0	1 (4.8) [1]
Permanent withdrawal of CZP due to TEAEs	0	0	0	0
TEAEs requiring dose change	0	1 (100) [1]	0	1 (4.8) [1]
Drug-related TEAEs	1 (6.7) [1]	0	0	1 (4.8) [1]
Discontinuation due to TEAEs	1 (6.7) [1]	0	0	1 (4.8) [1]
All deaths (TEAEs leading to death)g	0	0	0	0

^an was the number of participants who reported at least one TEAE in that category

^b[#] was the number of individual occurrences of the TEAE in that category

^cOne participant had their dose reduced from CZP 400 mg Q2W to CZP 200 mg Q2W. This participant was only counted once under the initial dose group

^dOne participant was included in the Not Dosed group; an event of “small for gestational age” was reported for this participant, which, upon further review, was determined to not be a true event by the investigator. This participant subsequently re-enrolled and was included in the CZP 400 mg Q4W group and the All Dosed Participants group

^eA serious adverse event was defined as any untoward medical occurrence that, at any dose if any of the following occurred: resulted in death; was life-threatening; required inpatient hospitalization or prolongation of existing hospitalization; resulted in persistent disability/incapacity; was a congenital anomaly/birth defect; was an important medical event

^fSevere TEAEs included one case of foetal death and one case of premature rupture of membranes

^gAll AEs were considered TEAEs as participants were already on CZP before the study began. Thus ‘All deaths’ were counted based on TEAEs leading to death

CZP certolizumab pegol, mg milligram, Q2W every 2 weeks, Q4W every 4 weeks, SS safety set, TEAE treatment-emergent adverse event

There were five (23.8%) incidences of serious TEAEs (Table 2), including three events related to pregnancy, puerperium and perinatal conditions. One serious TEAE of premature rupture of membranes occurred at gestation week 19 and led to study discontinuation (n = 1): this participant did not provide subsequent postpartum blood samples. One serious AE of foetal death (n = 1) occurred at gestation week 25 within a twin pregnancy in a mother with risk factors including obesity, a history of hypothyroidism and hypertension, in addition to the underlying inflammatory condition (RA). This AE was considered not treatment-related and did not lead to CZP interruption or study discontinuation. One serious AE of spontaneous abortion was reported in an enrolled participant before their first dose of CZP in the study (n = 1). Individual cases of spontaneous abortion/miscarriage and foetal death were analysed in detail, and it was concluded that there was no new safety information that altered the known benefit–risk balance of CZP.

Other serious TEAEs reported comprised one case each of hyponatraemia and haemorrhage (Table 2). None of the serious TEAEs were considered drug-related, as determined by the investigator.

Among all dosed participants, there was one drug-related TEAE of interest of basal cell carcinoma which was not considered serious, and which did not lead to treatment interruption or study discontinuation (n = 1). One other TEAE of interest of was reported (haemorrhage [System Organ Class: vascular disorders, i.e. not pregnancy, puerperium and perinatal conditions]; n = 1). No study participant deaths, infant illnesses, new safety signals or other safety concerns were reported.

Discussion

In this phase 1B study, we report that based on the statistical model of pooled data during pregnancy, relative to postpartum, there was a modest reduction in pre-dose plasma CZP concentrations that was consistent across trimesters 1 through 3, whilst there was no clear pattern in post-dose concentrations.

The PK of CZP has been previously investigated in multiple studies in different indications in non-pregnant individuals [21–23]. In this study, during pregnancy relative to postpartum, there was a reduction in plasma CZP concentrations that was consistent across trimesters 1 through 3; however, observed concentrations still broadly overlapped with the range seen postpartum. Additionally, the observed concentrations in the present study (in which patients were enrolled following ≥ 12 weeks of treatment with commercial CZP, at which time it was assumed that plasma CZP levels had reached a steady state) were within the range observed in CZP studies of non-pregnant individuals with PSO, axSpA and RA following 12 weeks of treatment with CZP 200 mg Q2W (mean week 12 plasma CZP concentrations: NCT04740814: 28.0 ug/ml, NCT03051217: 16.2 ug/ml, and NCT02505542: 31.7 ug/ml) [21–23]. When the linear mixed effects model was fitted to the pre-dose and post-dose CZP concentrations separately, pre-dose plasma CZP concentrations were consistently modestly lower throughout pregnancy compared to postpartum. However, there was no clear pattern in post-dose plasma CZP concentrations.

As expected for pregnancy, BMI increased and albumin decreased from trimester 1 to trimester 3; however, there was no consistent change in CRP over pregnancy (data not shown). Adjusting individually for the variables BMI and albumin, but not CRP, resulted in a smaller reduction of plasma CZP concentrations during pregnancy relative to postpartum in the linear mixed effects modelling, suggesting association of these variables with the observed difference in CZP concentrations during pregnancy.

The findings of the present study are supported by those of a recently published population PK analysis that also reported data from the CHERISH study, which demonstrated that although an increase in apparent clearance of CZP was associated with advancing through the stages of pregnancy and high anti-drug antibody titre, there was substantial overlap between simulated systemic CZP exposure in the pregnant and a non-pregnant population [24]. Hence, both the population PK analysis and the analysis in this manuscript support the maintenance of established CZP dosing regimens throughout pregnancy. [24]

Importantly, the safety profile observed in this study was in alignment with events expected over the course of pregnancy and was consistent with the known safety profile of CZP [17, 25]. One case of foetal death within a twin pregnancy was reported in the study; this was considered to be unrelated to treatment, and did not lead to CZP interruption or study discontinuation. No study participant deaths, infant illnesses, new safety signals or other safety concerns were reported. These results align with a previous analysis of exposure to TNFis in pregnant women which reported no teratogenic effect nor increased risk of foetal death associated with CZP treatment and further supports the safety of continuing established CZP dosing regimens during pregnancy. [26]

The formation of ADAbs, or immunogenicity, has been described in response to biologic therapies [27]. The use of a highly-sensitive MSD ECLIA assay in this study demonstrated rates of ADAb positivity consistent with the incidence seen in CZP studies of non-pregnant individuals with PsA, axSpA, and RA employing the same ADAb analysis method [22, 23]. In the present study, 85.7% of participants were ADAb positive at enrolment, with a range of 42.9–90.0% ADAb positivity throughout pregnancy (compared with a 100% [NCT03051217] and 96.1% [NCT02505542] overall incidence of ADAb positivity in prior studies using the same method). [22, 23]

Additionally, high and consistent rates of ADAb positivity were expected considering that participants in CHERISH were only eligible to enroll once they had been on a stable, maintenance dose of CZP for at least 12 weeks, at which point it was expected that steady state CZP levels had been reached.

Limitations

The interpretations of these data are limited by the small number of participants on different dosing regimens (400 mg Q2W and 400 mg Q4W treatment groups) and CIs that were difficult to meaningfully interpret. As a result, comparisons between dosing regimens could not be made. Furthermore, the models adjusting mean CZP concentrations based on BMI and serum albumin displayed higher variability compared to the model without them, which may limit the interpretation of the role of these parameters in the PK differences between pregnancy and postpartum. Additionally, this study did not examine how the observed changes in CZP plasma concentrations or the influence of ADAb concentration may have impacted disease activity. Finally, due to the highly sensitive CZP assay used in this analysis, it is not appropriate to compare absolute plasma concentrations reported here with those reported in previous studies employing different methodologies [28, 29], given the substantially greater sensitivity and specificity of the assay used in the present study.

Conclusions

Adequate disease control during pregnancy is crucial for women living with CIDs to ensure the best maternal and foetal health outcomes and reduce adverse pregnancy outcomes. However, to date, there has been limited knowledge on the PK of biologics and TNFi agents during pregnancy [30, 31]. In the present study, PK data of CZP in pregnant women with CIDs showed that CZP plasma concentrations were modestly lower during pregnancy versus postpartum but remained consistent throughout pregnancy. These findings are supported by those of a prior population PK analysis of data from the CHERISH study, which demonstrated a large overlap of the simulated CZP exposure between pregnant and non-pregnant individuals. The safety profile of CZP was consistent with the known safety profile of CZP in the general patient population and was in alignment with events expected over the course of pregnancy. Collectively these findings present an important contribution to inform appropriate dosing of CZP during pregnancy in women with CIDs, supporting the maintenance of established CZP dosing regimens in order to maintain stable plasma concentrations.

Supplementary Information

Below is the link to the electronic supplementary material.

Author Contributions

Substantial contributions to study conception and design: Megan E. B. Clowse, Radboud J. E. M. Dolhain, Stephanie Finzel, Frauke Förger, Cornelia Glaser, Andrea Pluma, Laura Shaughnessy, Jagdev Sidhu, Jemma Greenin, Kathy Rice, Gauri Utturkar, Joao N. Duarte, Marie Teil; substantial contributions to analysis and interpretation of the data: Megan E. B. Clowse, Radboud J. E. M. Dolhain, Stephanie Finzel, Frauke Förger, Cornelia Glaser, Andrea Pluma, Laura Shaughnessy, Jagdev Sidhu, Jemma Greenin, Kathy Rice, Gauri Utturkar, Joao N. Duarte, Marie Teil; drafting the article or revising it critically for important intellectual content: Megan E. B. Clowse, Radboud J. E. M. Dolhain, Stephanie Finzel, Frauke Förger, Cornelia Glaser, Andrea Pluma, Laura Shaughnessy, Jagdev Sidhu, Jemma Greenin, Kathy Rice, Gauri Utturkar, Joao N. Duarte, Marie Teil; final approval of the version of the article to be published: Megan E. B. Clowse, Radboud J. E. M. Dolhain, Stephanie Finzel, Frauke Förger, Cornelia Glaser, Andrea Pluma, Laura Shaughnessy, Jagdev Sidhu, Jemma Greenin, Kathy Rice, Gauri Utturkar, Joao N. Duarte, Marie Teil.

Funding

This study was sponsored by UCB. The journal’s Rapid Service Fee was funded by UCB.

Data Availability

Due to the small sample size in this trial, individual patient-level data cannot be adequately anonymised as there is a reasonable likelihood that individual participants could be re-identified. For this reason, data from this trial cannot be shared.

Declarations

Conflict of Interest

Megan E. B. Clowse: Grant/research support from: UCB, GSK. Consultant for: UCB, GSK; Radboud J. E. M. Dolhain: Unrestricted research grants from: Dutch Arthritis Association, ZonMw, UCB and Galapagos. Speaking fees from: UCB, Roche, AbbVie, Genzyme, Novartis, AstraZeneca and Eli Lilly; Stephanie Finzel: Grant/research support from: Novartis. Consultant for: Novartis, NovoNordisk. Speaker’s honoraria: AbbVie, Chugai, Galapagos, GSK, Novartis, UCB; Frauke Förger: Grant/research support from: UCB. Speaker’s bureau: Mepha, Roche, UCB, MSD; Cornelia Glaser: Speaker’s honoraria: Galapagos, GSK, Pfizer, UCB; Andrea Pluma: Speaker’s honoraria: UCB. Support for attending educational events: Fresenius, UCB. Participation Advisory Board: Galapagos; Laura Shaughnessy, Jagdev Sidhu, Jemma Greenin, Kathy Rice, Joao N. Duarte, Marie Teil: Employee and shareholder of UCB. Gauri Utturkar: Former Employee of UCB; employee of GSK.

Ethical Approval

The CHERISH study protocol, amendments, and patient informed consent were reviewed by a national, regional, or Independent Ethics Committee (IEC) or Institutional Review Board (IRB) (Supplementary Table 4). CHERISH was conducted in accordance with the current version of the applicable regulatory and International Conference on Harmonisation (ICH)-Good Clinical Practice (GCP) requirements, the ethical principles that have their origin in the principles of the Declaration of Helsinki, and the local laws of the countries involved. Full details of the ethics committee involved are available upon request. All patients provided written informed consent to participate in the study, with the addition of obtaining parental consent where relevant.