GP2017-HCF, a high concentration formulation, demonstrates similar pharmacokinetics, immunogenicity and safety to GP2017, an approved adalimumab biosimilar

ABSTRACT

Background

GP2017 is an adalimumab biosimilar. The objective of this study is to compare the pharmacokinetics (PK) of GP2017 in its approved formulation and GP2017-high concentration formulation (HCF) in a randomized, double-blind, two-arm PK bridging study.

Research design and methods

Healthy male subjects received a single 40 mg subcutaneous injection of either GP2017-HCF (n = 162) or GP2017 (n = 168). PK, safety, and immunogenicity were assessed over 72 days post-injection.

Results

The 90% confidence intervals [CIs] of geometric mean ratios between GP2017-HCF and GP2017 for C_max, AUC_0-inf, AUC_0-360 and AUC_0-last were within the pre-defined margin of 0.80 to 1.25; thus, PK comparability between GP2017-HCF and GP2017 was demonstrated. Subgroup analysis of PK comparability by anti-drug antibody (ADA) subpopulation showed that the 90% CIs of geometric mean ratios between GP2017-HCF and GP2017 for C_max, AUC_0-inf, AUC_0-360 and AUC_0-last were within the margin of 0.80 to 1.25 in ADA-positive and ADA-negative subjects. The proportions of subjects with positive ADA responses and with neutralizing antibodies were comparable between the GP2017-HCF and GP2017 groups. GP2017-HCF and GP2017 were well tolerated, and there were no reports of deaths or other serious adverse events.

Conclusion

Results show PK comparability between GP2017-HCF and GP2017 and comparable safety and tolerability.

KEYWORDS:

• Adalimumab
• antidrug antibody
• biosimilar
• GP2017
• high concentration formulation
• immunogenicity
• pharmacokinetics
• PK comparability

This article is related to:

A Plain Language Summary Describing how Two Different Concentrations of GP2017, a Biosimilar Adalimumab Medicine, are Associated with Similar Drug Levels within the Body

1. Introduction

Adalimumab is a human monoclonal antibody directed against tumor necrosis factor alpha (TNF-α), which is approved for and has significantly improved the treatment of several immune-mediated inflammatory diseases, including rheumatoid arthritis (RA), juvenile idiopathic arthritis, psoriatic arthritis (PsA), ankylosing spondylitis (AS), Crohn’s disease (CD), ulcerative colitis (UC), plaque psoriasis (PsO), hidradenitis suppurativa (HS), and uveitis [1]. Sandoz Biopharmaceuticals developed adalimumab-adaz, GP2017 (Hyrimoz®) as a biosimilar to the adalimumab reference medicine (Humira®, AbbVie Inc., United States). GP2017 has been approved by the Food and Drug Administration (FDA) [2,3] and European Medicines Agency (EMA) [4] in 2018 as a safe and efficacious alternative to its reference medicine. GP2017 is administered subcutaneously (s.c.) using a prefilled syringe (PFS) or an autoinjector/prefilled pen (AI).

GP2017 has been shown to have identical amino acid sequences, indistinguishable secondary and tertiary structures and the same level of post-translational modifications as its reference medicine [5]. Preclinical data showed that GP2017 matches adalimumab reference medicine in its biological functionality, including similar target binding, functional, and pharmacodynamic properties [6]. Clinically, GP2017 has demonstrated similar pharmacokinetics (PK) to the reference medicine following administration of single doses to healthy volunteers, with comparable PK following administration with PFS or AI [7]. Furthermore, GP2017 has shown similar efficacy, safety, and immunogenicity to its reference medicine in two Phase 3 studies in patients with moderate-to-severe PsO [8] and in patients with moderate-to-severe RA [9]. Treatment with GP2017 resulted in comparable improvements in patient-reported outcomes (PROs) and quality of life scores in patients with PsO, PsA, and RA. Switching between GP2017 and its reference medicine had no negative impact on PROs [10]. Biosimilars aim to decrease cost, thereby increasing patient access; data from a nationwide study conducted in Denmark provided evidence for cost savings of 82.8% when switching all patients from the reference medicine to a biosimilar adalimumab [11].

GP2017 was initially developed in a formulation containing adalimumab-adaz at a concentration of 50 mg/mL, enabling administration of a 40 mg dose per 0.8 mL injection. Sandoz Biopharmaceuticals has developed a high concentration formulation (GP2017-HCF) increasing the drug concentration to 100 mg/mL, reducing the injection volume and thereby increasing patient convenience. A dose of 80 mg in 0.8 mL can now be administered in a single injection, decreasing the number of injections for the induction treatment for CD, UC, PsO, HS, or uveitis or for dose escalations during maintenance treatment, if required. GP2017-HCF was developed based on the approved GP2017 and filed to health authorities in 2022. Both formulations contain the same active ingredient, adalimumab-adaz. In contrast to GP2017, GP2017-HCF does not contain citric acid and sodium chloride. The remainder of the excipients of GP2017-HCF are the same as those of GP2017, albeit at different concentrations.

Here, we present the results of the randomized, double-blind, parallel, two-arm PK bridging study, comparing the PK of GP2017 and GP2017-HCF, following administration of a single dose of 40 mg to healthy volunteers. In accordance with the International Conference on Harmonization Q5E guideline (ICH Q5E), which provides guidance on the comparability assessment of biotechnological/biological products subject to changes in their manufacturing process [12], the two formulations were compared following similar principles as required for the introduction of an altered formulation of a reference biological medicine.

2. Patients and methods

2.1. Study design and subjects

This was a multicenter, randomized, double-blind, parallel, two-arm, single-dose study to evaluate the PK, safety, and immunogenicity of GP2017-HCF (high-concentration formulation; 100 mg/mL) and GP2017 (50 mg/mL) in healthy adult male subjects. The study consisted of three parts: screening, randomization and treatment, and the follow-up period. Subjects were eligible if they were 18–55 years old, weighed 65–110 kg, with a body mass index (BMI) ranging between 20 and 29.9 kg/m², had no previous exposure to adalimumab (adalimumab reference medicine or its biosimilars) and no exposure to other biologics in a clinical trial setting during the 9 months prior to screening. Subjects who participated in any clinical investigation with small molecules within two months or five half-lives or within nine months if a therapeutic protein (e.g. a monoclonal antibody or fusion protein) was administered were excluded from the study. Subjects with quantifiable anti-TNF serum concentration or ADAs against adalimumab at screening were excluded from the study. Also subjects with active infections or inflammatory conditions within two weeks prior to screening were excluded. All eligible subjects were randomized (1:1) to one of the two treatment groups (GP2017-HCF or GP2017). Body weight is a known covariate impacting the PK of adalimumab [13,14]. Therefore, during randomization, subjects were stratified by body weight categories (65.0 to <76.0 kg, 76.0 to <92.0 kg, and 92.0 to 110.0 kg) with the aim of obtaining a balanced number of subjects between the two treatment groups per weight category. Subjects were also stratified by study site to obtain a balanced number of subjects between the two treatment groups at each study site.

The subjects received a single 40 mg s.c. injection of either GP2017-HCF or GP2017 on Day 1. Thereafter, assessments and follow-up were conducted from Day 2 to Day 72 (end-of-study visit). Subjects could voluntarily withdraw from the study for any reason at any time. Subjects who dropped out of the study after randomization were not replaced. The study was conducted in accordance with the ICH E6 Guideline for Good Clinical Practice, and with the ethical principles outlined in the Declaration of Helsinki. All subjects provided written informed consent. The study protocols were approved by the Institutional Review Board (Advarra, 6940 Columbia Gateway Dr. Suite 110, Columbia, MD, USA).

2.2. Study objectives and endpoints

The primary objective of the study was to demonstrate PK comparability between GP2017-HCF and GP2017 in healthy male subjects. The primary endpoints were maximum serum concentration (C_max), area under the concentration–time curve measured from the time of dosing and extrapolated to infinity (AUC_0-inf), and area under the concentration–time curve measured from the time of dosing to 360 hours post-dose (AUC_0-360). The key PK secondary endpoints were area under the concentration–time curve measured from the time of dosing to the last measurable sampling timepoint (AUC_0-last), percentage of AUC_0-inf extrapolated from the time of last observed concentration to infinity (%AUC_extrap), elimination half-life (t_1/2), time to reach maximum serum drug concentration (t_max), and slope of the serum concentration curve in the terminal elimination phase (lambda_z). Secondary objectives also included the assessment of immunogenicity, safety, and local tolerability of GP2017-HCF and GP2017. Local tolerance in terms of injection-site reactions (ISRs) was assessed by an ISR score, and injection site pain (ISP) was assessed by a pain visual analogue scale (VAS) immediately after the injection and at 15 min and 4 h post injection on Day 1. For the ISP assessment, VAS-scores (range 0 mm – 100 mm) were categorized as no pain (<5 mm), mild pain (≥5 to <45 mm), moderate pain (≥45 to <75 mm), and severe pain (≥75 mm) [15,16].

2.3. Pharmacokinetic assessments

Blood samples (5 mL) were collected from a forearm vein (either by direct venipuncture or from an indwelling catheter) into vacutainer tubes at each time point (pre-dose and 6, 9, 24, 48, 72, 96, 120, 144, 168, 192, 216, 360, 528, 696, 1032, 1368, and 1704 hours post-dose) and serum was separated. All serum samples were collected, labeled, processed, and stored frozen at ≤−70°C until analysis. Serum concentration measurements were performed at bioanalytical laboratories (Hexal AG, Oberhaching, Germany).

A validated enzyme-linked immunosorbent assay (ELISA)-based method was used to determine the drug concentration in serum samples. The adalimumab target, TNF-α, was used for capturing, while an enzyme-labeled antibody was used for detection. The assay linear range was defined within lower (LLOQ) and upper limit of quantitation (ULOQ) of 250 ng/mL and 8000 ng/mL, respectively; samples with concentrations above ULOQ were diluted appropriately. Quality control samples had a validated inter-assay accuracy of 95–99% and a precision expressed as coefficient of variation of 5–9%. Method validations were performed according to international regulatory guidelines for bioanalytical methods. The concentrations of drug in the serum and quality control samples were determined by interpolation using a standard curve based on known concentrations of adalimumab-adaz (Sandoz, Germany). Concentrations were expressed in mass per volume units (ng/mL). Mean, median and range were calculated for each nominal timepoint. Values below the LLOQ (BLQ) were considered zero for summary statistics. If the BLQ occurred between two quantifiable drug concentrations within a full PK profile, BLQ value was considered missing. C_max, t_1/2 and t_max were obtained from the serum concentration of the subjects. The PK parameters including AUC_0-inf, AUC_0-360, AUC_0-last and percentage of AUC_0-inf (%AUC_extrap) were calculated for each subject by non-compartmental analysis (using Phoenix WinNonlin Version 8.1).

2.4. Safety and immunogenicity assessments

Safety assessments consisted of collecting data on all adverse events (AEs) and serious AEs (SAEs), including their severity and relationship to study treatment. Other parameters relevant for the safety assessment included physical examinations, vital signs, body weight, laboratory investigations (clinical chemistry, hematology, and urinalysis), electrocardiograms, and local tolerance.

Anti-drug antibody (ADA) formation before and after the study drug administration, and further evaluation of their potential neutralizing effect were analyzed by immunoassays. Samples for detection of ADA were collected pre-dose and on Days 3, 7, 10, 16, 23, 30, 44, 58, and 72 post-dose. Immunogenicity was determined by testing for binding and neutralizing antibodies against the study drug using validated immunoassays. The immune response after the study drug administration was evaluated by a multi-tiered approach including a validated bridging immunogenicity assay used for the screening, confirmation, and titration of binding ADAs. A validated ELISA-based competitive ligand-binding assay (CLB assay) was used for assessing the neutralizing capacity of the antibodies (i.e. the capacity to interfere with binding of study drug to its target). Screening visit samples were only analyzed in the ADA screening and confirmatory assay. The false-positive rates for the screening, confirmatory, and titer ADA assays were calculated to be 5%, 1%, and 0.1%, respectively. The false-positive rate of the neutralizing antibody (NAb) assay was calculated to be 1%.

All samples were first analyzed in a screening assay. Study samples with a result below the validated screening cut-point were considered negative for binding ADAs and were reported accordingly. All samples with a result above or equal to the screening assay cut-point were additionally analyzed in a secondary confirmatory assay (specificity assay). If the result of the confirmatory assay was above or equal to the validated confirmatory cut-point, the sample was reported as confirmed positive for binding anti-adalimumab antibodies. The titers of all confirmed ADA-positive samples were determined in a subsequent analysis . In addition, all confirmed positive ADA samples were analyzed in the CLB assay to evaluate the neutralizing potential of the adalimumab-specific antibodies. Study samples with a result below the validated NAb assay cut-point were considered NAb negative while samples with a result above or equal to the NAb assay cut-point were considered NAb positive.

As high levels of adalimumab in the serum samples can interfere with the ADA screening assay, the drug tolerance of this assay was evaluated during the assay method validation. In addition, adalimumab concentrations were determined for each ADA sample to exclude a potential interference with adalimumab in serum samples. The assay sensitivity was determined to be 58 ng/mL and the drug tolerance for the assay was 50 µg/mL, that is, above the highest drug concentration measured in the clinical study.

2.5. Statistical analysis

The randomized set, which included all subjects who were randomized, including those who did not receive study treatment, was used for the analyses of subject disposition only. Subjects were analyzed according to randomization. The safety analysis set, which included all subjects who received the study treatment, was used for all analyses of safety, immunogenicity, and demographic data. Subjects were analyzed according to the study treatment received.

The PK analysis set (PKS) was used for the analysis of PK parameters. The PKS included all subjects who received the study treatment and completed the study without a major protocol deviation that had a relevant impact on PK data. Subjects with positive reverse transcription–polymerase chain reaction (RT-PCR) SARS-CoV-2 test results during the study (Day 1 to Day 72) were excluded. Subjects were analyzed according to the study treatment received. In addition to all subjects with major protocol deviations, 15 subjects were excluded from the analysis of AUC_0-inf, and other lambda_z dependent PK parameters, because less than 3 data points in the terminal phase were available for these subjects or subjects had abrogated adalimumab serum concentration–time profiles (Table S1).

The primary endpoint assessment of PK comparability was based upon the 90% confidence intervals (CIs) of the geometric mean ratios of GP2017-HCF and GP2017 for C_max, AUC_0-inf and AUC_0-360. AUC_0-inf was included as a co-primary endpoint for the assessment of PK comparability as per FDA recommendation, whereas EMA agreed on the assessment of PK exposure using AUC_0-360 as a sensitive co-primary endpoint which is less influenced by the development of ADA [17]. To conclude PK comparability, the 90% CIs of the geometric mean ratios for primary endpoints had to be contained within the PK comparability limits of 0.80 to 1.25. This method is equivalent to conducting 2 one-sided tests at the 5% significance level [18].

An analysis of covariance (ANCOVA) was performed separately on each of the ln-transformed PK parameters C_max, AUC_0-inf and AUC_0-360. The ANCOVA model included treatment as a fixed effect and baseline body weight (at Day 1) as a covariate. The ANCOVA calculated the least squares (LS) means for the treatments; the ratios of LS mean (GP2017-HCF/GP2017) were calculated using the exponentiation of the difference in LS means from the analyses on the corresponding ln-transformed PK parameters. For each PK parameter, the 90% CIs for these ratios were derived.

3. Results

3.1. Subject disposition and demographics

Of 777 subjects screened, 331 were randomized. One subject in the GP2017-HCF group was randomized by mistake and withdrawn from the study before receiving study treatment. A total of 330 subjects received study treatment (GP2017-HCF: 162; GP2017: 168) and were included in the analysis, of which 315 subjects (95.5%) completed the study. A total of 16 subjects out of 331 randomized subjects (4.8%) discontinued the study. Demographics and baseline characteristics, including body weight and BMI, were well balanced between the two treatment groups (Table 1).

Table 1. Subject demographics (Safety set).

Characteristic	GP2017-HCF N = 162	GP2017 N = 168
Age (years), mean (SD)	36.8 (9.04)	37.8 (9.52)
Male, n (%)	162 (100)	168 (100)
Race, n (%)
American Indian or Alaska Native	0	1 (0.6)
Asian	1 (0.6)	1 (0.6)
Black or African American	38 (23.5)	26 (15.5)
Multiple	1 (0.6)	0
White	122 (75.3)	140 (83.3)
Ethnicity, n (%)
Hispanic or Latino	134 (82.7)	140 (83.3)
Not Hispanic or Latino	28 (17.3)	28 (16.7)
Height (cm), mean (SD)	173.7 (6.47)	174.4 (6.65)
Weight (kg), mean (SD)	81.21 (9.10)	81.24 (8.79)
Weight stratification factor, n (%)
65.0 – <76.0 kg	50 (30.9)	50 (29.8)
76.0 – <92.0 kg	94 (58.0)	100 (59.5)
92.0–110.0 kg	18 (11.1)	18 (10.7)
BMI (kg/m^2), mean (SD)	26.90 (2.24)	26.70 (2.34)

BMI, body mass index; SD, standard deviation

3.2. Pharmacokinetics analysis

The PKS comprised 300 subjects (GP2017-HCF: 152; GP2017: 148), after exclusion of 31 subjects (18 subjects were excluded for major protocol deviations, 12 subjects did not complete the study or had a positive SARS-CoV-2, and 1 subject was randomized by mistake and discontinued the study before receiving study treatment), Table S1. Mean serum concentration-time profiles of GP2017-HCF and GP2017 are shown in Figure 1. The mean adalimumab serum concentration–time profiles were comparable between both groups.

Figure 1. Mean (±SD) linear (a) and semilogarithmic (b) serum concentration-time profiles (Safety analysis set).

The PK comparisons of GP2017-HCF and GP2017 are presented in Table 2 and Table S2. The 90% CIs of geometric mean ratios of GP2017-HCF and GP2017 for C_max, AUC_0-inf, AUC_0-360 and AUC_0-last were within the pre-defined margin of 0.80 to 1.25 (Table 2 and Figure 2). Thus, PK comparability between GP2017-HCF and GP2017 was demonstrated.

Figure 2. Forest plot of PK endpoints (PK analysis set).

Table 2. Comparison of the PK parameters of GP2017-HCF and GP2017 (PK analysis set).

PK parameter (unit)	Treatment	n	Geometric LS means	Ratio GP2017-HCF/ GP2017
				Point estimate	90% CI(Lower, Upper)
Cmax (ng/mL)	GP2017-HCF	152	3247	1.0242	(0.9536, 1.1001)
	GP2017	148	3170
AUC0-inf (h*ng/mL)	GP2017-HCF	143	2334410	1.0605	(0.9758, 1.1525)
	GP2017	142	2201267
AUC0-360 (h*ng/mL)	GP2017-HCF	152	904669	1.0318	(0.9568, 1.1127)
	GP2017	148	876775
AUC0-last (h*ng/mL)	GP2017-HCF	152	1859656	1.0432	(0.9404, 1.1572)
	GP2017	148	1782651

n = number of subjects by treatment group with evaluable PK parameter data. ANCOVA was performed for each PK parameter including treatment as a fixed effect and baseline body weight (at Day −1) as a covariate. ANCOVA, analysis of covariance; AUC, area under the concentration–time curve; C_max, maximum serum concentration; PK, pharmacokinetics

Subgroup analysis of PK comparability by ADA subpopulation showed that the 90% CIs of geometric mean ratios of GP2017-HCF and GP2017 for C_max, AUC_0-inf, AUC_0-360 and AUC_0-last were within the margin of 0.80 to 1.25 in ADA-positive and ADA-negative subjects (Table 3). These results underline the PK comparability of GP2017-HCF and GP2017 in both subpopulations.

Table 3. PK comparability by ADA subpopulation (PK analysis set).

				Ratio GP2017-HCF/ GP2017
PK parameter (unit)	Treatment	n	Geometric LS means	Point estimate	90% CI(Lower, Upper)
ADA-positive
Cmax (ng/mL)	GP2017-HCF	106	3246	1.0505	(0.9597, 1.1499)
	GP2017	98	3090
AUC0-inf (h*ng/mL)	GP2017-HCF	98	2087731	1.1085	(1.0022, 1.2260)
	GP2017	92	1883389
AUC0-360 (h*ng/mL)	GP2017-HCF	106	896459	1.0751	(0.9761, 1.1841)
	GP2017	98	833858
AUC0-last (h*ng/mL)	GP2017-HCF	106	1625243	1.0918	(0.9574, 1.2451)
	GP2017	98	1488565
ADA-negative
Cmax (ng/mL)	GP2017-HCF	46	3252	0.9759	(0.8676, 1.0976)
	GP2017	50	3332
AUC0-inf (h*ng/mL)	GP2017-HCF	45	2970576	1.0108	(0.9097, 1.1232)
	GP2017	50	2938790
AUC0-360 (h*ng/mL)	GP2017-HCF	46	926803	0.9608	(0.8546, 1.0803)
	GP2017	50	964593
AUC0-last (h*ng/mL)	GP2017-HCF	46	2536196	0.9990	(0.8982, 1.1111)
	GP2017	50	2538700

n = number of subjects by treatment group with evaluable PK parameter data. ANCOVA was performed for each PK parameter including treatment as a fixed-effect and baseline body weight (at Day −1) as a covariate. ANCOVA, analysis of covariance; AUC, area under the concentration–time curve; C_max, maximum serum concentration; PK, pharmacokinetics

3.3. Immunogenicity

The proportions of subjects with positive ADA responses and with NAbs were comparable between the GP2017-HCF and GP2017 groups, overall as well as at all individual visits (Table 4). Of note, two subjects in the GP2017-HCF and four subjects in the GP2017 group, were tested ADA-positive at Day 7, that is, during the latency period of 10 days that is characteristic for a primary immune response [19].

Table 4. Summary of immunogenicity results (Safety analysis set).

Visit	Result	GP2017-HCFN = 162 n (%)	GP2017N = 168 n (%)
Day 1 (Pre-dose)	ADA Positive	1 (0.6)	0
	NAb Positive	1 (0.6)	0
Day 3	ADA Positive	0	0
	NAb Positive	0	0
Day 7	ADA Positive	2 (1.2)	4 (2.4)
	NAb Positive	2 (1.2)	4 (2.4)
Day 10	ADA Positive	2 (1.2)	5 (3.0)
	NAb Positive	2 (1.2)	4 (2.4)
Day 16	ADA Positive	58 (35.8)	50 (29.8)
	NAb Positive	26 (16.0)	24 (14.3)
Day 23	ADA Positive	68 (42.0)	61 (36.3)
	NAb Positive	34 (21.0)	40 (23.8)
Day 30	ADA Positive	43 (26.5)	47 (28.0)
	NAb Positive	29 (17.9)	36 (21.4)
Day 44	ADA Positive	41 (25.3)	44 (26.2)
	NAb Positive	40 (24.7)	44 (26.2)
Day 58	ADA Positive	68 (42.0)	72 (42.9)
	NAb Positive	67 (41.4)	72 (42.9)
Day 72	ADA Positive	94 (58.0)	90 (53.6)
	NAb Positive	93 (57.4)	90 (53.6)
Overall (up to Day 72)	ADA Positive	112 (69.1)	109 (64.9)
	ADA Negative	49 (30.2)	59 (35.1)
	NAb Positive	102 (63.0)	103 (61.3)
	Transient	18 (11.1)	17 (10.1)

The number of subjects of each category at each visit (n) is displayed in the table. Percentage is calculated as 100*n/N. ‘Overall’ indicates a subject had an ADA negative test result at baseline and had at least 1 ADA positive result (ADA-positive) or consistently negative results (ADA-negative) post-baseline. ‘Transient’ indicates that a subject experienced a final negative ADA result at any visit following a positive ADA result.

ADA, anti-drug antibody; NAb, neutralizing antibody

3.4. Safety and tolerability

Single-dose administrations of 40 mg s.c. GP2017-HCF and GP2017 were safe and well tolerated. There were no reports of deaths or other SAEs. The number of subjects who discontinued due to treatment-emergent AEs (TEAEs) was small and comparable between the two treatment groups (GP2017-HCF: 1 subject; GP2017: 3 subjects). The overall proportions of subjects with TEAEs were comparable between the GP2017-HCF (80 subjects, 49.4%) and the GP2017 (95 subjects, 56.5%) treatment groups (Table 5). The overall safety profile in this study was comparable between both treatments and in accordance with the established safety profile of GP2017 [20].

Table 5. TEAEs by primary system organ class and preferred term (at least 2% of subjects in any treatment group) (Safety set).

Primary system organ class Preferred term	GP2017-HCFN = 162n (%)	GP2017N = 168n (%)
Number of subjects with at least 1 event	80 (49.4)	95 (56.5)
General disorders and administration site conditions	55 (34.0)	69 (41.1)
Injection site reaction	50 (30.9)	63 (37.5)
Investigations	16 (9.9)	24 (14.3)
Hemoglobin decreased	4 (2.5)	5 (3.0)
Alanine aminotransferase increased	5 (3.1)	3 (1.8)
Aspartate aminotransferase increased	4 (2.5)	4 (2.4)
Weight decreased	2 (1.2)	6 (3.6)
Musculoskeletal and connective tissue disorders	8 (4.9)	14 (8.3)
Back pain	3 (1.9)	4 (2.4)
Pain in extremity	1 (0.6)	6 (3.6)
Nervous system disorders	8 (4.9)	6 (3.6)
Headache	6 (3.7)	4 (2.4)
Skin and subcutaneous tissue disorders	10 (6.2)	4 (2.4)
Rash	4 (2.5)	1 (0.6)
Gastrointestinal disorders	8 (4.9)	3 (1.8)
Respiratory, thoracic and mediastinal disorders	2 (1.2)	6 (3.6)
Eye disorders	1 (0.6)	4 (2.4)
Injury, poisoning and procedural complications	4 (2.5)	1 (0.6)

TEAEs are events that started or worsened after study treatment. A subject with multiple occurrences of an event within the same primary system organ class or preferred term was counted only once. Primary SOCs and PTs are sorted in descending frequency (total column of source table). MedDRA version 24.1

PT, preferred term; SOC, system organ class; TEAEs, treatment-emergent adverse events

Overall, 114 subjects (34.5%) reported ISRs during the study. Numerical differences observed in the proportions of subjects reporting at least 1 ISR between the GP2017-HCF (51 subjects, 31.5%) and the GP2017 groups (63 subjects, 37.5%) were not considered clinically meaningful. All ISRs reported during the study were mild in severity, and none of the ISRs were reported as serious or required concomitant medication except for the event of vessel puncture pain. ISP severity, as assessed by VAS, was comparable between the 2 treatment groups at all time points. (Figure 3/ Table S3).

Figure 3. Box plot of VAS scores.

4. Discussion

Comparative PK bridging studies form an important part of the comparability assessment, given that PK endpoints are typically more sensitive than efficacy endpoints [21]. The mechanisms involved in absorption of large therapeutic proteins following s.c. administration are not fully understood [22,23]. In addition, there is limited understanding regarding the influence of the formulation and the physicochemical characteristics of excipients on s.c. absorption of large therapeutic proteins. Therefore, the development of GP2017-HCF, the high concentration (100 mg/mL) formulation of the adalimumab biosimilar GP2017, was supported by a clinical PK bridging study following administration of single doses of 40 mg to healthy volunteers in accordance with ICH Q5E.

PK bridging studies determining whether a complex formulation change does impact PK, as well as PK similarity studies determining the similar PK of a proposed biosimilar to its reference medicine, are ideally conducted in healthy subjects following administration of single doses [24]. Healthy subjects represent a sensitive population to detect potential differences in PK since they do not receive concomitant medication and have no underlying disease that could impact the PK study endpoints. Furthermore, healthy subjects are immune-competent, and unlike patients, are devoid of confounding effects impacting their immune response [25], and thus represent a sensitive population to exclude differences in immunogenicity triggered by a manufacturing change.

The subjects of this PK bridging study were selected based on stringent inclusion criteria minimizing inter-subject variability. This further enhances the sensitivity to detect potential differences in PK resulting from the tested manufacturing changes by removing confounding, non-product related variability. Changes in water content of plasma during the menstrual cycle may lead to the higher variability of exposure to drugs predominantly distributing in plasma, including adalimumab and other monoclonal antibodies that has been observed in woman compared to men [26]. Only healthy male volunteers were therefore included in this study, this approach is generally accepted. There are no known gender-specific differences in PK, efficacy, or safety of GP2017 [27].

Adalimumab induces ADAs in a high proportion of patients and healthy subjects. The reference product as well as its biosimilars are known to be highly immunogenic, single-dose administration in healthy volunteers lead to ADA in 44–98% of subjects [28,29]. In general, GP2017 PK similarity studies, including the PK bridging study reported here, were carried out in a parallel group design. A cross-over design is not feasible as this would potentially lead to an impact of earlier ADA formation on later PK assessments. A population PK analysis of adalimumab in healthy subjects has shown that body weight and the presence of ADAs impact the PK of adalimumab following the administration of single doses [30]. Therefore, the allocation to the two treatment groups in this study was stratified by body weight groups. To minimize the impact of ADAs on GP2017 PK, which has been shown to comprise elements of inter-individual variability in treatment-naive healthy subjects [17], a screening for prior recent anti-TNFα therapeutic antibody exposure as well as for preexisting ADAs was integrated into the screening phase of the study. In addition, early ADA sampling on Days 3, 7 and 10 post-dose was introduced, to detect possible secondary immune responses to adalimumab which occur at timepoints earlier than 10–14 days post-dose and with higher titers. This was enabled by the high drug-tolerance of the ADA-assay used in this study which exhibited a drug tolerance of 50 µg/mL (i.e. above the reported C_max) following the administration of a single 40 mg dose.

Non-product related differences in immunogenicity increase the variability of PK parameters [17]. AUC_0-360 is a measure of exposure less influenced by ADA development and is therefore a more sensitive and robust endpoint than AUC_0-inf for highly immunogenic drugs. We proposed establishing PK comparability between GP2017 and GP2017-HCF based on AUC_0-360 to the FDA and EMA. While the EMA accepted this proposal, the FDA requested AUC_0-inf, the standard primary endpoint following single-dose administration, as a co-primary endpoint.

To minimize the likelihood of a false-positive outcome of this PK bridging study, a total of 330 participants were planned to be randomized to achieve an overall power of 90% to test comparability of the co-primary endpoints C_max and AUC_0-inf between the two GP2017 formulations (assuming a coefficient of variation of 42.2% for AUC_0-inf and 30.0% for C_max). Given the lower variability for AUC_0-360 in comparison to AUC_0-inf, this sample size provided an overall power of 98% to test comparability of the co-primary PK endpoints C_max and AUC_0-360 (assuming a coefficient of variation of 30.6% for AUC_0-360).

The results of the PK bridging study showed that the new GP2017-HCF matches the approved GP2017 formulation in all tested PK parameters. C_max is a sensitive endpoint to detect potential differences in s.c. absorption. All measures of GP2017 exposure, i.e., AUC_0-360 and AUC_0-inf represent sensitive endpoints to detect potential differences in elimination. The 90% CIs for the geometric mean ratios of all these endpoints were within the prespecified margin (0.80–1.25), demonstrating PK comparability. In addition, the statistical comparison of the key secondary PK endpoint AUC_0-last matched between the two treatment arms. For all comparisons, the point estimates were close to unity and the 90% CIs contained 1. All other PK parameters, including the secondary PK parameters %AUC_extrap, t_l/2, t_max, and lambda_z, were comparable between GP2017 and GP2017-HCF.

Immunogenicity, the rate and amount of ADA developing over time, following administration of a single dose of GP2017, was measured with an ADA-assay tolerant to drug concentrations above C_max. The numbers and proportions of subjects with positive ADA and NAb responses were comparable between the two treatment groups, overall as well as at all individual visits. The pre-specified subgroup analysis of PK in ADA-positive and ADA-negative subjects was within the comparability margin of 0.80–1.25, providing further evidence for comparable immunogenicity of the two GP2017 formulations.

For five subjects, two in the GP2017-HCF group and three in the GP2017 treatment group, adalimumab serum concentration–time profiles were abrogated, that is, adalimumab serum concentrations were BLQ before 360 hours after dosing in these subjects. Four of the five subjects had an early onset of ADA formation on Day 7 post-dose, and ADA titers were >1000 in these four subjects, reached at Day 10 post-dose. The early onset of ADAs, as well as the high ADA titers in these subjects are indicative of a secondary immune response. To confirm this hypothesis, immunoglobulin isotyping was performed on these ADA samples and the results will be reported elsewhere. Further exploratory analyses to identify factors responsible for differences in inter-individual immune responses to GP2017 were conducted in this study. The results of these analyses will also be reported elsewhere.

The incidence and severity of ISR were comparable in the GP2017-HCF and GP2017 treatment groups. Moreover, all ISRs were of mild intensity.

ISP following s.c. administration of a medicine is influenced by multiple factors, such as formulation, volume of injection, injection process, injection devices, and several patient-related factors [31,32]. There are conflicting results regarding the local tolerance of adalimumab formulations containing citrate buffer, thereby creating uncertainty among health-care professionals as well as patients [33,34]. Citrate buffer is commonly used in parenteral formulations at concentrations in the range of 5–15 mM and it is assumed that increasing its concentration to more than 50 mM can result in excessive pain upon s.c. injection [31]. GP2017 has a low citrate concentration (1.2 mM). Volume can also influence the pain perception: injection of larger volumes (≥1.2 mL) can be associated with higher perception of pain, while no difference in injection site pain at lower injection volumes has been demonstrated [35]. This study, therefore, investigated the impact of the low citrate concentration of GP2017, compared to the new citrate-free GP2017-HCF, by means of a VAS score assessing ISP.

At all timepoints, the VAS scores were similar in both treatment groups and median VAS scores were below 5 mm and thus categorized as ‘no pain’ in line with interpretation of VAS ratings and change scores [15] (Table S3). The highest VAS scores observed were similar for both formulations (Figure 3). The numerical differences in median VAS score at the timepoint of the most sensitive ISP assessment, that is, immediately after s.c. administration (3.0 mm for GP2017-HCF vs 4.0 mm for GP2017) were below the minimum clinically meaningful difference of 12 mm [36]. Moreover, GP2017-HCF showed comparable ISP to corresponding data obtained in other HCF adalimumab [37,38].

5. Conclusions

The results reported here show the PK comparability of GP2017-HCF, a newly developed, high-concentration, low volume, citrate-free formulation of GP2017 demonstrating equal bioavailability of the two formulations. The safety profile and tolerability were comparable between both the treatments.

Declaration of interest

O von Richter, D Guerrieri, C Fey, F Furlan, and L Lemke are employees of Hexal AG/Sandoz. J Fan and S Schussler are employees of Sandoz Inc. T O’Reilly is an employee of Celerion Inc. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

Reviewer disclosures

Peer reviewers on this manuscript have received an honorarium from Expert Opinion on Biological Therapy for their review work. A reviewer on this manuscript has disclosed that they are working on a study with AbbVie which they do not believe results in any relevant conflicts of interest. Peer reviewers on this manuscript have no other relevant financial relationships or otherwise to disclose.

Author contributions

O von Richter and L Lemke have been involved in the conception and design, analysis and interpretation of the data and the drafting of the paper. T O’Reilly, D Guerrieri, and S Schussler have been involved the conception and design, in the analysis and interpretation of the data. C Fey, J Fan, and F Furlan have been involved in the analysis and interpretation of the data. All authors have critically revised the paper for intellectual content, have approved the final version of the manuscript and agree to be accountable for all aspects of the work.

Ethics approval

The study was conducted in accordance with the ICH E6 Guideline for Good Clinical Practice, and with the ethical principles outlined in the Declaration of Helsinki. All subjects provided written informed consent. The study protocols were approved by the Institutional Review Board (Advarra, 6940 Columbia Gateway Dr. Suite 110, Columbia, MD, USA).

Acknowledgments

The authors thank Rajeeb Ghosh of Novartis Healthcare for providing medical writing assistance in accordance with Good Publication Practice (GPP3) guidelines, which was funded by Hexal AG/Sandoz.

Supplementary material

Supplemental data for this article can be accessed online at https://doi.org/10.1080/14712598.2022.2117546

Additional information

Funding

This study was sponsored by Hexal AG/Sandoz.