Abstract
Aims
To compare the pharmacokinetics, safety, tolerability and immunogenicity of FKB327, a biosimilar of adalimumab, with European Union (EU)‐approved Humira and US‐licensed Humira after single subcutaneous doses in healthy subjects.
Methods
In a randomized, double‐blind, parallel‐group study, 180 healthy subjects received by subcutaneous injection 40 mg of EU‐Humira, or US‐Humira, or FKB327, in a 1:1:1 ratio, stratified by bodyweight. Pharmacokinetics, local tolerability, immunogenicity, adverse events, vital signs, electrocardiography and laboratory safety tests were assessed prior to and up to 1536 h after treatment.
Results
The pharmacokinetics of FKB327 were similar to those of both EU‐ and US‐Humira. The 90% confidence interval for the ratios of AUC0–t, AUC0–inf, and Cmax geometric means were in the acceptance range for bioequivalence of 0.80–1.25 for all three pairwise comparisons by analysis of covariance with baseline characteristics age, body weight and (for Cmax only) sex as covariates. Tolerability of all three treatments was equally acceptable, and there were no differences in safety profile or immunogenicity among the three treatments. Overall, antidrug antibodies were detected in approximately 70% of subjects who received each treatment; higher titres were associated with faster elimination of adalimumab.
Conclusions
The study demonstrated pharmacokinetic similarity of FKB327 with EU‐ and US‐Humira. FKB327 was well tolerated by healthy subjects, with adverse effects similar to Humira. If clinical similarity to Humira, including efficacy, can be shown in patients, FKB327 will meet the criteria for biosimilarity to Humira.
Keywords: adalimumab, biosimilar, FKB327, Humira, monoclonal, TNF‐α
What is Already Known About this Subject
Humira (adalimumab) is a therapeutic monoclonal antibody targeted against human tumour necrosis factor‐α, and is widely used for the treatment of autoimmune diseases.
What this Study Adds
FKB327 is a new biological product that contains the monoclonal antibody adalimumab as its active ingredient, making it a potential biosimilar product to Humira.
FKB327 showed similar pharmacokinetics to EU‐Humira and US‐Humira in healthy subjects. Safety, tolerability and immunogenicity were also similar.
EU‐Humira and US‐Humira had similar pharmacokinetics in healthy subjects.
Tables of Links
| LIGANDS |
|---|
| Adalimumab |
These Tables list key protein targets and ligands in this article that are hyperlinked to corresponding entries in http://www.guidetopharmacology.org, the common portal for data from the IUPHAR/BPS Guide to PHARMACOLOGY 1, and are permanently archived in the Concise Guide to PHARMACOLOGY 2015/16 2.
Introduction
Adalimumab is a recombinant human monoclonal antibody against human tumour necrosis factor‐α (TNF‐α). Adalimumab was first approved for the treatment of rheumatoid arthritis in December 2002 in the USA 3, 4 and in September 2003 in the European Union (EU) 5, and was subsequently launched globally under the brand name Humira. Humira has proved to be a safe and effective treatment for patients with various autoimmune diseases 6. Protection of the active ingredient by patents will end shortly, which will allow the marketing of competing products that contain the same monoclonal antibody as the reference drug 7, 8, 9, 10, 11, provided that other patents (such as those covering manufacture and formulation) are not infringed. Such competing products are known as ‘biosimilar’ medicines. Generic drugs of chemical origin have structurally identical active substances to the lead product; they generally require demonstrated equivalence of only pharmacokinetics to gain marketing approval. However, for biosimilar drugs, the potential structural differences in the active substance, which result from the biological manufacturing process, require additional equivalence studies prior to marketing, including efficacy studies in patients 11.
FKB327 is being developed as a medicine that may prove to be biosimilar to Humira. FKB327, like Humira, contains adalimumab that is expressed in and purified from Chinese hamster ovary cells. The strength of FKB327 is the same as that of the Humira dosage form that was originally approved (40 mg adalimumab per 0.8 mL deliverable solution). Both FKB327 and Humira are formulated such that their pH is around pH 5.2, but they contain different excipients, buffers and stabilizer.
We recently completed a study of the first administration of FKB327 to healthy human subjects. The pharmacokinetics, safety, tolerability and immunogenicity of FKB327 were compared with those of EU‐approved Humira (EU‐Humira) and US‐licensed Humira (US‐Humira). This study is the first step in the clinical development process required to obtain marketing approval of this biosimilar product.
Methods
This study was done at Hammersmith Medicines Research, a clinical pharmacology unit in the UK, in accordance with the prevailing EU directives and International Conference on Harmonisation Good Clinical Practice. After review of the safety profile of adalimumab, and the safety record of monoclonal antibody studies in healthy subjects, we considered it acceptable to do the study 6, 12, 13, 14, 15. The study was not started until it had received approval from the Medicines and Healthcare products Regulatory Agency (MHRA; EudraCT number: 2012–005140‐23), and Scotland A Research Ethics Committee (reference number: 12/SS/0220). Subjects gave written, informed consent. The trial was conducted between April and November 2013.
Objectives
The primary objective was to compare the safety and pharmacokinetics of FKB327, EU‐Humira and US‐Humira after single doses, by subcutaneous injection, in healthy humans.
The secondary objective was to assess immunogenicity and tolerability.
Subjects
To be eligible for the study, subjects had to be of age 18–65 years, weight 60–90 kg, and body mass index 18–30 kg m–2. Subjects were excluded if they had previously received adalimumab, but were not excluded if they had received a different therapeutic monoclonal antibody.
There were 180 healthy men and women enrolled in the study. All women were of nonchildbearing potential. All subjects gave written consent after being informed about the study and its potential risks. They were deemed healthy after assessment of medical history and a medical examination that included vital signs, electrocardiography (ECG) and routine laboratory safety tests. Subjects were also tested for drugs of abuse and pregnancy (women only) prior to the start of the study. During the study, strenuous exercise, alcohol, caffeine and concomitant medication (other than paracetamol) were restricted. Smoking was not allowed during periods of residence and for 24 h prior to visits; at other times during the study, up to five cigarettes daily were allowed.
Design
This was a randomized, double‐blind, parallel‐group study in healthy subjects, to compare the pharmacokinetics, safety, tolerability and immunogenicity of single doses of subcutaneously administered FKB327 with Humira from two regions. Subjects were assigned to receive 40 mg of either EU‐Humira, US‐Humira or FKB327, in a 1:1:1 ratio (60 subjects to receive each treatment).
As FKB327 had not previously been administered to humans, the first 12 subjects to be dosed were split into three subgroups of three, three and six participants, with at least 7 days between dosing of the first two subgroups; the final subgroup was dosed after at least a further 2 days. The three treatments were randomized in a 1:1:1 ratio within each subgroup. The remaining subjects were randomized in a 1:1:1 ratio stratified by weight (using two strata of ≤75 kg and >75 kg).
Subjects were screened within 4 weeks prior to their dose of adalimumab. Subjects were resident in the ward from the morning prior to dosing until 8 days afterwards. They returned for outpatient visits about 15, 22, 29, 36, 43, 50 and 64 days after their dose.
Assessments
Blood samples for pharmacokinetic analysis were taken prior to dosing and at 4, 12, 24, 36, 48, 72, 96, 120, 144, 168, 192, 360, 528, 696, 864, 1032, 1200 and 1536 h afterwards. Serum concentrations of adalimumab were determined using a validated immunoassay on an electrochemiluminescent (ECL) platform using 96‐well Meso Scale Discovery (MSD) high‐bind plates coated with TNF‐α. The lower limit of reliable quantification was 100 ng ml–1.
Safety assessments were done prior to dosing and regularly until the subject's final visit. Assessments included: vital signs (heart rate, blood pressure, respiration rate, and oral temperature), 12‐lead ECG, and laboratory safety tests of blood and urine. In addition, ECG telemetry was done for 4 h after dosing.
Tolerability assessments included: local tolerability checks at the injection site immediately after dosing and at 12, 24, 48, 72, 96 and 192 h after dosing, and visual analogue scales for pain (0 = no pain; 100 = intolerable pain) immediately after dosing and at 0.5, 1, 12 and 24 h after dosing, using standard methods described elsewhere 16, 17.
Blood samples for assessment of immunogenicity were taken prior to dosing and at 360, 696 and 1536 h afterwards. Immunogenicity [i.e. presence of anti‐drug antibodies (ADA) to adalimumab] was determined by a validated immunoassay on an ECL platform using MSD plates to detect ADA against adalimumab in the form of FKB327, EU‐Humira, and US‐Humira. The confirmation of specificity of the three assays for FKB327, EU‐Humira and US‐Humira used a floating cut‐off point of 28.5–35.7% inhibition. Titres were classified as: negative; below the lower limit of reliable quantification; 1, 4, 16, 64, 256, 1024, 4096 or 16 384; or above the upper limit of reliable quantification.
The neutralizing capability of ADA was assessed using MSD plates on an ECL platform to assess the ability of serum samples to block the binding of ruthenylated adalimumab to TNF‐α immobilized on the assay plate. When applicable, neutralizing ADA results for all three investigational drugs were reported with annotations stating that the circulating drug concentrations exceeded the drug tolerance limits of the assays (> 500 ng ml–1). Samples analysed in the presence of drug concentrations above the tolerance limits of the assays made the results inconclusive. Samples that tested negative in the ECL immunoassay were not tested for neutralizing ADA.
Sample size and statistical methods
Sample size calculations were done using nQuery Advisor 7.0, assuming that the pharmacokinetic parameters AUC0–t (area under concentration–time curve up to last nonzero value), AUC0–∞ (area under concentration–time curve extrapolated to infinity) and Cmax (peak serum concentration) were the primary endpoints. The coefficient of variation of pharmacokinetic parameters was predicted to be 35% on the basis of data obtained in healthy volunteer studies and published by the US Food & Drug Administration 18. In total, 180 subjects (60 per treatment arm) had to complete the study to achieve 94% power so that the 90% confidence intervals (CI) for the ratio of geometric means fell within the acceptance criterion of (0.80, 1.25), for each FKB327:Humira comparison, and for the EU‐Humira:US‐Humira comparison, assuming no difference existed. As there were three treatment comparisons, the overall power was lower than 94%, but not significantly so, as the three equivalence tests are highly correlated.
Pharmacokinetic data underwent formal statistical analysis, to test whether the main parameters met the equivalence criteria. The primary hypothesis was that FKB327 had similar pharmacokinetics to both EU‐Humira and US‐Humira, and that EU‐Humira and US‐Humira were similar to one another, based on equivalence analysis of the geometric means of AUC0–t, AUC0–∞ and Cmax. The secondary hypothesis was equivalence of AUC0–360h (area under concentration–time curve up to 360 h) and t½ (elimination half‐life). The pharmacokinetic parameters were logarithmically transformed prior to analysis by analysis of covariance (ANCOVA) which included a term for treatment group. Covariates (age, body weight, body surface area and sex) were retained in the model only if significant at the 10% level. The least squares geometric means were weighted using the average of the demographic terms included in the model.
All other data were summarized using standard descriptive statistics.
Results
Subject disposition and demography
There were 180 subjects (170 men, and 10 women of nonchildbearing potential) enrolled in the study, who received a single subcutaneous injection of FKB327, EU‐Humira or US‐Humira (60 subjects received each treatment). Demographic data are presented in Table 1. Overall, subjects were aged 18–64 years, and had a body mass index within the range 18.7–29.4 kg/m2, and weighed 60.1–89.4 kg. Most subjects (65%) self‐reported as being of White ethnicity; the remaining subjects claimed Black (17.2%), Asian (12.8%) or Other (5.0%) ethnicity. The demographic characteristics of subjects were similar across treatment groups.
Table 1.
Summary of demographic data for each treatment group
| FKB327 (N = 60) | EU‐Humira (N = 60) | US‐Humira (N = 60) | Total (N = 180) | |
|---|---|---|---|---|
| Age (years) | ||||
| Mean (SD) | 31.0 (10.95) | 35.2 (14.08) | 32.3 (12.35) | 32.8 (12.57) |
| Median (range) | 28.0 (19–64) | 29.5 (18–64) | 28.0 (19–62) | 29.0 (18–64) |
| Sex, n (%) | ||||
| Male | 58 (96.7) | 55 (91.7) | 57 (95.0) | 170 (94.4) |
| Female | 2 (3.3) | 5 (8.3) | 3 (5.0) | 10 (5.6) |
| Race, n (%) | ||||
| Asian | 6 (10.0) | 5 (8.3) | 12 (20.0) | 23 (12.8) |
| Black or African American | 14 (23.3) | 9 (15.0) | 8 (13.3) | 31 (17.2) |
| White | 34 (56.7) | 45 (75.0) | 38 (63.3) | 117 (65.0) |
| Other | 6 (10.0) | 1 (1.7) | 2 (3.3) | 9 (5.0) |
| BMI at screening (kg m–2) | ||||
| Mean (SD) | 24.01 (2.281) | 23.75 (2.297) | 24.24 (2.750) | 24.00 (2.447) |
| Range | 19.8–28.6 | 19.4–29.4 | 18.7–29.3 | 18.7–29.4 |
| Weight at screening (kg) | ||||
| Mean (SD) | 74.59 (7.924) | 74.87 (7.466) | 73.67 (8.352) | 74.38 (7.895) |
| Range | 61.2–89.3 | 61.8–89.4 | 60.1–88.9 | 60.1–89.4 |
N, total number of subjects; n, number of subjects per group; BMI, body mass index; SD, standard deviation.
All subjects completed the study, apart from one in the FKB327 treatment group, who withdrew himself from the study after Day 29, and missed the final four outpatient visits.
Pharmacokinetics
All 180 subjects were included in the pharmacokinetic analysis.
The serum concentration–time profiles of the three adalimumab treatments showed kinetics typical of subcutaneously delivered monoclonal antibodies: serum concentrations peaked several days after dosing (median tmax 6–8 days) and t½ was about 2 weeks. The serum concentration–time profiles of the three treatments were similar (Figure 1 and Table S1), and were consistent with previous observations 4, 5. Pharmacokinetic parameters are summarized in Table 2.
Figure 1.
Linear (top) and semi‐log (bottom) mean serum concentration–time plots of adalimumab after a 40 mg dose for FKB327, EU‐Humira and US‐Humira
Table 2.
Primary and secondary pharmacokinetic parameters of adalimumab for FKB327, EU‐Humira and US‐Humira
| Pharmacokinetic parameter | Geometric mean (geometric CV) | ||
|---|---|---|---|
| FKB327 N = 60 | EU‐Humira N = 60 | US‐Humira N = 60 | |
| Cmax (ng ml–1) | 3310 (34.5) | 2830 (32.0) | 3100 (34.2) |
| tmax (h) | 144 (36.0–364)a | 192 (72.0–865)a | 144 (48.0–504)a |
| AUC0–360h (h*ng ml–1) | 958 000 (35.0) | 824 000 (33.7) | 918 000 (34.6) |
| AUC0–t (h*ng ml–1) | 2 130 000 (32.9) | 1 930 000 (35.9) | 2 130 000 (41.3) |
| AUC0–∞ (h*ng ml–1) | 2 350 000 (35.8)b | 2 170 000 (38.8)c | 2 410 000 (45.8)c |
| λz (/h) | 0.00214 (47.4)b | 0.00201 (48.4)c | 0.00189 (49.3)c |
| t½ (h) | 324 (47.4)b | 345 (48.4)c | 366 (49.3)c |
CV, coefficient of variation; N, total number of subjects; n, total number of subjects per group; Cmax, peak plasma concentration; tmax, time of peak plasma concentration; AUC0–360h, area under concentration–time curve up to 360 h; AUC0–t, area under concentration–time curve up to last nonzero value; AUC0–∞, area under concentration–time curve extrapolated to infinity; λz, apparent terminal elimination rate constant; t½, elimination half‐life.
median (range)
n = 58
n = 59
Data from all 180 subjects were included in the ANCOVA. The significant covariates age and body weight were included, but body surface area, which is correlated to body weight, was not significant and was therefore removed from the model. Sex was significant for Cmax only, so was not included in the model for AUC0–∞ or AUC0–t. For the secondary parameters, the covariates were forced to be the same as for the primary AUC parameters. The averages of the demographic terms used for weighting of the least squares geometric means were: weight 74.38 kg, age 32.8 years and (for Cmax only) male:female ratio 170/10.
For each of the primary parameters (AUC0–∞, AUC0–t, Cmax) the 90% CI for the ratio of geometric least squares means from the ANCOVA analysis (Table 3) were within the predefined limits of 0.80–1.25 for all treatment comparisons. Furthermore, the 90% CI for the ratio of all primary parameters encompassed 1.00, with the exception of Cmax in the FKB327:EU‐Humira comparison, where the lower 90% CI was just above 1.00. Based on the ANCOVA, similarity was concluded among the three treatments: FKB327, EU‐Humira and US‐Humira. For the secondary parameters, AUC0–360h was equivalent in all three treatment comparisons; t½ was equivalent for two treatment comparisons, but not FKB327:US‐Humira, because the 90% CI extended just below 0.80.
Table 3.
Summary of ANCOVA of primary and secondary adalimumab pharmacokinetic parameters: ratio of geometric least squares means with 90% confidence intervals (CI)
| Pharmacokinetic parameter | Ratio of geometric least squares means (90% CI) | |||
|---|---|---|---|---|
| FKB327/ EU‐Humira | FKB327/ US‐Humira | EU‐Humira/ US‐Humira | ||
| Primary | AUC0–∞ (h*ng ml–1) | 1.06 (0.94, 1.18)a | 0.98 (0.88, 1.10)a | 0.93 (0.83, 1.04)a |
| AUC0–t (h*ng ml–1) | 1.08 (0.97, 1.20)a | 1.01 (0.91, 1.12)a | 0.93 (0.84, 1.03)a | |
| Cmax (ng ml–1) | 1.13 (1.03, 1.23)a | 1.07 (0.98, 1.17)a | 0.95 (0.87, 1.04)a | |
| Secondary | AUC0–360h (h*ng ml–1) | 1.12 (1.02, 1.23)a | 1.04 (0.95, 1.14)a | 0.93 (0.85, 1.02)a |
| t½ (h) | 0.95 (0.83, 1.10)a | 0.90 (0.78, 1.03) | 0.94 (0.82, 1.08)a | |
AUC0–∞, area under concentration–time curve extrapolated to infinity; AUC0–t, area under concentration–time curve up to last nonzero value; Cmax, peak plasma concentration; AUC0–360h, area under concentration–time curve up to 360 h; t½, elimination half‐life.
Note: for Cmax, age, weight and sex were included in the model; for both AUC0–∞ and AUC0–t, age and weight were included in the model. For the secondary parameters, the covariates were forced to be age and weight as for the primary AUC parameters.
90% CI within predefined limits (0.80, 1.25) concluding equivalence
Safety and tolerability
One hundred and ten subjects (61.1%) had at least one adverse event after dosing, but most events were only mild or moderate in severity; 97 subjects (53.9%) experienced adverse events that were considered to be related to adalimumab. Of those, three subjects (1.7%; one in each treatment group) had severe adverse events after dosing: head injury (EU‐Humira); dizziness (US‐Humira); and loss of consciousness (FKB327). The most common adverse events were headache, upper respiratory tract infection, nasopharyngitis, oropharyngeal pain and injection site haematoma (Table 4). The adverse event profile was similar across the three treatment groups (Table 4 and Table S2).
Table 4.
Treatment‐emergent adverse events reported by ≥5% of subjects after any treatment
| FKB327 N = 60 n (%) | EU‐Humira N = 60 n (%) | US‐Humira N = 60 n (%) | Total N = 180 n (%) | |
|---|---|---|---|---|
| Subjects with at least 1 treatment‐emergent adverse event | 35 (58.3) | 39 (65.0) | 36 (60.0) | 110 (61.1) |
| Headache a | 12 (20.0) | 11 (18.3) | 10 (16.7) | 33 (18.3) |
| Upper respiratory tract infection a | 5 (8.3) | 6 (10.0) | 4 (6.7) | 15 (8.3) |
| Oropharyngeal pain | 4 (6.7) | 2 (3.3) | 3 (5.0) | 9 (5.0) |
| Injection site haematoma a | 4 (6.7) | 1 (1.7) | 2 (3.3) | 7 (3.9) |
| Nasopharyngitis | 2 (3.3) | 2 (3.3) | 3 (5.0) | 7 (3.9) |
| Rash a | 1 (1.7) | 3 (5.0) | 2 (3.3) | 6 (3.3) |
| Back pain a | 0 | 3 (5.0) | 2 (3.3) | 5 (2.8) |
| Gastroenteritis | 0 | 2 (3.3) | 3 (5.0) | 5 (2.8) |
| Liver function test abnormal a | 2 (3.3) | 0 | 3 (5.0) | 5 (2.8) |
| Neutropenia | 3 (5.0) | 0 | 2 (3.3) | 5 (2.8) |
| Pain in extremity | 1 (1.7) | 3 (5.0) | 1 (1.7) | 5 (2.8) |
| Dizziness | 0 | 3 (5.0) | 1 (1.7) | 4 (2.2) |
| Nausea a | 0 | 3 (5.0) | 1 (1.7) | 4 (2.2) |
N, total number of subjects; n, total number of subjects per group. Treatment‐emergent adverse events are those that started or increased in severity after the dosing. Events were coded using Medical Dictionary for Regulatory Activities (MedDRA) version 16.1. Each subject is counted only once within each preferred term. Events are ordered per overall preferred term frequency.
An adverse reaction reported by ≥ 5% of patients in placebo‐controlled clinical trials of Humira, which occurred more frequently than placebo
Two subjects had serious adverse events, both of which were considered to be possibly related to adalimumab. One subject in the FKB327 treatment group had transient loss of consciousness (initially reported as seizure), which resolved fully: no definitive diagnosis could be made. One subject in the US‐Humira group developed a psychotic disorder, which required admission to hospital and treatment with antipsychotic drugs: resolution was only partial.
Nine subjects had laboratory results of potential clinical importance, after dosing. Of those, low neutrophil count was most common (but was deemed to be related to subjects' race 19), followed by deranged liver function tests. None of the laboratory results was of clinical concern: although mean neutrophil count decreased slightly from 24 h post dose to Day 30 in all treatment groups, the changes in neutrophil count did not differ among treatment groups. Results of other safety assessments were similar across treatment groups, and no result was of clinical concern.
Injection site pain
Most subjects had no irritation at the injection site (Table 5). Immediately after dosing, 10 subjects (5.6%) had minimal erythema: most of those were in the EU‐(five subjects) and US‐(four subjects) Humira treatment groups, and only one was in the FKB327 treatment group. All local tolerability assessments were normal from 12 h after dosing onwards.
Table 5.
Visual analogue scales for pain at the injection site after each adalimumab treatment
| Time point after injection (h) | Mean [median] (range) pain score | ||
|---|---|---|---|
| FKB327 N = 60 | EU‐Humira N = 60 | US‐Humira N = 60 | |
| Immediately after dosing | 5.5 [1.5] (0–54) | 12.9 [9.5] (0–48) | 18.4 [11.0] (0–86) |
| 0.5 | 2.1 [0.0] (0–29) | 1.4 [0.0] (0–19) | 2.1 [0.0] (0–35) |
| 1 | 1.4 [0.0] (0–25) | 0.7 [0.0] (0–15) | 0.5 [0.0] (0–6) |
| 12 | 0.4 [0.0] (0–6) | 0.4 [0.0] (0–13) | 0.5 [0.0] (0–20) |
| 24 | 0.1 [0.0] (0–2) | 0.3 [0.0] (0–8) | 0.1 [0.0] (0–2) |
N, total number of subjects
Immediately after dosing, subjects had higher scores for injection site pain in the EU‐Humira (median 9.5; range 0–48) and US‐Humira (median 11.0; range 0–86) treatment groups than did subjects in the FKB327 treatment group (median 1.5; range 0–54), although scores varied widely among subjects (Table 5). At all other time points, pain scores were similar across treatment groups, with median score of 0, ranging from 0–35 at 0.5 h after dosing to 0–8 at 24 h after dosing.
Immunogenicity and effect on pharmacokinetics
About 5% of subjects had detectable ADA prior to dosing. The proportion of subjects with positive ADA at the last sampling point (1536 h after dosing), when drug concentrations were least likely to interfere with the assay, was similar among treatments: 69.5% for FKB327, 73.3% for EU‐Humira, and 70.0% for US‐Humira (Table 6). The development of ADA specific to FKB327, EU Humira and US‐Humira was also similar among treatments, as was the proportion of subjects at each level of ADA titre. In addition, cross reactivity between the three ADA assays was high (i.e. subjects who tested positive for ADA ostensibly specific to their adalimumab treatment also tested positive for ADA specific to the other two adalimumab treatments).
Table 6.
Frequency of negative and positive tests for antidrug antibodies (ADA, specific to treatment administered) after each adalimumab treatment
| Planned time (h) | FKB327 N = 60 n (%) | EU‐Humira N = 60 n (%) | US‐Humira N = 60 n (%) | |||
|---|---|---|---|---|---|---|
| ADA result a | Negative | Positive | Negative | Positive | Negative | Positive |
| Predose | 57 (95.0) | 3 (5.0) | 57 (95.0) | 3 (5.0) | 57 (95.0) | 3 (5.0) |
| 360 | 39 (65.0) | 21 (35.0) | 41 (68.3) | 19 (31.7) | 45 (75.0) | 15 (25.0) |
| 696 | 39 (66.1) | 20 (33.9) | 41 (68.3) | 19 (31.7) | 42 (70.0) | 18 (30.0) |
| 1536 | 18 (30.5) | 41 (69.5) | 16 (26.7) | 44 (73.3) | 18 (30.0) | 42 (70.0) |
N, total number of subjects; n, number of subjects per group
Based on confirmatory results only; specific to treatment administered
Of the subjects who tested positive for ADA at the last sampling time point, about 83% also tested positive in the neutralizing antibody assay, with a further 10% having inconclusive results owing to interference by adalimumab at concentrations above >500 ng ml–1. ADA‐positive samples that yielded negative or inconclusive results in the neutralizing assay were mainly at low titres of 4 or less. Overall, the proportion of negative, inconclusive and positive neutralizing antibodies was similar among treatments.
The effect of ADA activity on adalimumab pharmacokinetics was further investigated. Mean serum concentrations of adalimumab in subjects with low ADA activity (lower quartile: ADA negative), moderate ADA activity (median: ADA titre <256), or high ADA activity (upper quartile: ADA titre ≥256) are presented for each treatment in Figure 2. Subjects with higher ADA activity showed faster elimination of adalimumab in all treatment groups. To investigate this further, scatter plots of Cmax, AUC0–∞ and t½ in individual subjects vs. ADA titre at the last sampling timepoint (1536 h, or sooner if missing) were prepared and presented separately for each treatment (Figure S1). The plots indicated that, for both AUC0–∞ and t½, high ADA titres were associated with lower AUC0–∞ and shorter t½. However, there did not appear to be a clear relationship between individual Cmax values and ADA titre.
Figure 2.
Mean serum concentration–time profiles of adalimumab by antidrug antibody activity in FKB327, EU‐Humira and US‐Humira treated subjects
Discussion
Biosimilarity can be concluded if biological products meet the standard criteria for bioequivalence with respect to their primary pharmacokinetic parameters, and if their safety, tolerability and efficacy are shown to be equivalent 9, 10. In this study, we sought to confirm in healthy subjects that FKB327 was pharmacokinetically similar to Humira products from two regions. We used a single subcutaneous injection of 40 mg adalimumab because that is the standard therapeutic dose for patients with rheumatoid arthritis, and the risks of administering a single 40 mg dose to healthy subjects were considered acceptable 4, 5, 12. After 40 mg FKB327, exposure to adalimumab was similar to that after Humira from both the EU and US. ANCOVA of the primary parameters (AUC0–∞, AUC0–t, Cmax) demonstrated pharmacokinetic similarity of FKB327 to both Humira products.
The ANCOVA accounts for between‐subject variability caused by demographic characteristics such as body weight, and estimates differences at the common mean level that may reduce bias where covariates are not perfectly balanced. ANCOVA was the pre‐specified analysis, and was included in the statistical analysis plan for the study. However, the analysis would have been more conservative had the selected covariates been prespecified based on prior knowledge rather than selected by a step‐wise approach. Posthoc analysis of variance (ANOVA) was also done to investigate the robustness of the results. Briefly, basic ANOVA, without baseline covariates, was done assuming equal variance of treatments. The results of the additional analysis are presented in Table S3: point estimates are broadly similar to the results of ANCOVA, but the confidence intervals are wider – as might be expected, given that the ANOVA did not take account of demographic variables. However, the analysis of primary pharmacokinetic parameters for the comparison FKB327:US‐Humira and EU‐Humira:US‐Humira confirmed pharmacokinetic similarity, as the 90% CI were still contained within the equivalence range 0.80–1.25.
For the primary parameter Cmax in the FKB327:EU‐Humira comparison, the 90% CI derived by ANOVA was wider than that from the corresponding ANCOVA, and the point estimate for the ratio of geometric least squares means was 1.17 by ANOVA vs. 1.13 by ANCOVA. Because of the different point estimates, for Cmax the upper 90% CI of 1.29 fell just outside the equivalence range (0.80–1.25) in the ANOVA but not in the ANCOVA. However, the other primary parameters, AUC0‐∞ and AUC0–t, met the equivalence criteria in the ANOVA.
The difference in outcome between ANCOVA and ANOVA analyses was predictable, because ANCOVA included as covariates several important demographic characteristics and so better accounted for differences among treatment groups than did ANOVA. Although the treatment groups were quite well balanced with respect to demography, there were nevertheless differences among groups that might plausibly account for the difference in outcome between ANCOVA and ANOVA, as follows. The EU‐Humira group had the highest proportion of women, and the FKB327 group the lowest, and subjects in the EU‐Humira group were slightly older, and more variable in age, than the FKB327 group; by accounting for those differences, ANCOVA yielded point estimates closer to unity than did ANOVA. Furthermore, body weight was slightly higher in the EU‐Humira group than in the other two groups. Body weight was a significant covariate in the ANCOVA: including it as a covariate yielded narrower confidence intervals than did ANOVA. Indeed, body weight is known to affect the pharmacokinetics of adalimumab 5, so it was used as a stratification factor in the randomization.
The differences among subjects in important demographic characteristics are accounted for in the ANCOVA model, so it is a more appropriate analysis than ANOVA for assessing pharmacokinetic similarity of the adalimumab formulations in this study. For assessing bioequivalence, both the EMA and FDA prefer ANOVA over ANCOVA unless the covariates selected for the model are predefined 20, 21. However, biosimilarity need not be precluded provided that any differences are analysed and concluded not to be clinically meaningful. In this study, the pharmacokinetic differences can be explained by intersubject variability arising from demographic characteristics as well as ADA activity.
Humira is manufactured in several different facilities. EU‐ and US‐Humira have the same excipients, but it is conceivable that there might be subtle differences in manufacture that could account for minor differences in bioavailability. However, the quality profile of Humira is well controlled 22. Several other biosimilars of Humira are in development, most of which have used Humira from the US and EU as reference materials 23, 24, 25, 26. Pharmacokinetic data have now been published for several products, (e.g. 23, 26): the results confirm our finding of pharmacokinetic similarity between EU‐ and US‐Humira. In September 2016, the US Food and Drug Administration approved Amjevita, a biosimilar to Humira, for multiple inflammatory diseases 27.
Although >60% of subjects in this trial reported at least one adverse event after treatment, most of the events were minor. Also, many of the events, such as headache, upper respiratory infection, and gastrointestinal disturbance, are common in untreated healthy people: a placebo treatment would be needed to judge whether their frequency in our subjects was higher than might have been expected. The prescribing information for Humira quotes the most common side effects as injection site reactions, respiratory tract infection, headache, abdominal pain, nausea and vomiting, rash and musculoskeletal pain, and most of those occurred in at least one subject in our study 4, 5. There were no important differences among treatments as regards nature of adverse events, but there was some variation in incidence. Injection site haematoma occurred in more subjects after FKB327 (four subjects) than after either of the Humira treatments (one and two subjects). Skin rash was less common after FKB327 (one subject) than after the Humira treatments (two and three subjects), but for both injection site disorders and skin rash the incidence was too low to allow any conclusions to be drawn. FKB327 and Humira contain the same active ingredient, but they differ in their excipients, buffers and method of manufacture. Those differences might affect local tolerability, which is why we assessed pain at the injection site. Injection site pain score for FKB327 immediately after dosing was lower than that for both Humira products, whereas the reverse pattern was seen for irritation at the injection site. Although injection site pain score varied widely among subjects, and significance testing would be unjustified, it is possible that injection site pain differed between FKB327 and Humira as result of their excipients. The EU recently approved a new formulation of Humira, whose local tolerability had been improved by the removal of certain excipients 28.
Overall, the side‐effect profile of adalimumab in healthy subjects seems similar to that seen in patients with autoimmune disease: the only unexpected events were one occurrence of loss of consciousness, and one episode of psychosis, neither of which seems likely to have been related to adalimumab treatment.
As expected with any formulation of adalimumab, there were no clinically important trends in vital signs or ECG, and there were few laboratory values of potential clinical importance. Some subjects had decreased neutrophil count 5–8 days after dosing, and two subjects had transient increases in serum transaminases from 5 days after dosing. Although those changes could have been caused by adalimumab, they could equally have been due to other factors such as race and lifestyle.
ADA were detected in up to 10–25% of patients receiving Humira without methotrexate 5, 13. In our study in healthy subjects, most treated subjects formed detectable ADA, with similar results in each treatment group. The difference between the present and previous studies is probably due to the high sensitivity of the assays we used in our study, although a true difference between healthy subjects and patients cannot be excluded and could be investigated further. We used a competitive ligand binding assay to detect neutralizing ADA. Ligand binding assays generally perform as well as do bioassays in the detection of antibodies, and, in some cases, may offer better sensitivity, lower variability, and less matrix interference 29. However, competitive ligand binding assays have the limitation that circulating drug can cause false positive results. The validation study of our ligand‐binding assay showed that drug concentrations of 1000 ng ml–1, but not 500 ng ml–1, in the samples could cause false positive results in the neutralizing ADA assay. In our clinical study, out of 118 neutralizing antibody positive samples at Day 65, only 13 had drug concentrations >500 ng ml–1. Thus, most of the neutralizing antibody positive results at Day 65 are likely to be truly ADA positive.
The prevalence of ADA to adalimumab varies with population and with analytical method. The highest reported prevalence of ADA is 87% (13 of 15 patients 30); in 11 of those 13 patients the drug was discontinued, suggesting that the ADA might have reduced or abolished the efficacy of adalimumab. Taking all the available information into account, we feel that the high prevalence of ADA in our study reflects mainly true ADA positive results of the ligand binding assay. It has also been reported that the pharmacokinetics of biosimilar products are affected by the presence of ADA 26. Guidelines on biosimilar products state that a pharmacokinetic comparability study should evaluate similarity not only in terms of absorption and bioavailability, but also in terms of possible differences between products in their elimination characteristics 9. Our data show a clear relationship between ADA titre and rate of adalimumab elimination. However, the prevalence of ADA positivity was similar among treatment groups. In particular, there was no clear difference in the distribution of ADA titre between FKB327 and Humira.
Overall, there was no difference in immunogenicity among the three treatment arms, and the proportion of negative, inconclusive and positive neutralizing ADA assay results was similar across treatments. In this study, high ADA titres were associated with a shorter t½ and a lower AUC0–∞ of adalimumab after each treatment. Cmax was unaffected by ADA titre, probably because after single doses, ADA do not appear until long after Cmax has been achieved. However, Cmax might well be reduced by ADA after the second and subsequent doses.
In conclusion, FKB327 was well tolerated by the healthy volunteers in this study, with a safety profile similar to that of Humira. The pharmacokinetics of FKB327 were similar to those of Humira from both the EU and US, and we judge the formulations to have similar pharmacokinetics. ADA accelerated the elimination of adalimumab from the body, but the prevalence of ADA and the elimination half‐life of adalimumab were similar after the three drug formulations. To confirm biosimilarity, a study is now required to show that FKB327 has clinically similar safety, tolerability, immunogenicity and efficacy to Humira in patients.
Competing Interests
The clinical part of this study, and data management, were done at Hammersmith Medicines Research Ltd. Statistical and pharmacokinetic analyses were done by Quanticate Ltd. Yasumasa Arai and Hideaki Nomura are employees of Fujifilm Kyowa Kirin Biologics Co, Ltd. No competing interests were declared from any other author.
This study was supported by Fujifilm Kyowa Kirin Biologics Co, Ltd.
The authors thank the volunteers and staff who participated in this study. We are grateful to Sue McKendrick (Quanticate Ltd) for reviewing the manuscript for its faithful presentation of the statistical analyses.
Data from this study have not previously been presented in any form.
Supporting information
Figure S1 Scatterplots of pharmacokinetic parameters vs. anti‐drug antibody (ADA) activity presented by treatment group
Table S1 Serum concentrations of adalimumab after a 40‐mg dose for FKB327, EU‐Humira and US‐Humira
Table S2 Treatment‐emergent adverse events reported by <5% of subjects in any treatment group (system organ class and preferred term)
Table S3 Summary of ANOVA of primary and secondary adalimumab pharmacokinetic parameters: ratio of geometric least squares means with 90% confidence intervals (CI)
Puri, A. , Niewiarowski, A. , Arai, Y. , Nomura, H. , Baird, M. , Dalrymple, I. , Warrington, S. , and Boyce, M. (2017) Pharmacokinetics, safety, tolerability and immunogenicity of FKB327, a new biosimilar medicine of adalimumab/Humira, in healthy subjects. Br J Clin Pharmacol, 83: 1405–1415. doi: 10.1111/bcp.13245.
References
- 1. Southan C, Sharman JL, Benson HE, Faccenda E, Pawson AJ, Alexander SP, et al. The IUPHAR/BPS guide to PHARMACOLOGY in 2016: towards curated quantitative interactions between 1300 protein targets and 6000 ligands. Nucleic Acids Res 2016; 44 (Database Issue): D1054–D1068. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Alexander SPH, Kelly E, Marrion N, Peters JA, Benson HE, Faccenda E, et al. The concise guide to PHARMACOLOGY 2015/16: Overview. Br J Pharmacol 2015; 172: 5729–5743. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Hussar DA. New drugs of 2003. J Am Pharm Assoc (2003) 2004; 44: 168–206 quiz 7‐10. [DOI] [PubMed] [Google Scholar]
- 4. FDA . Prescribing information – HUMIRA (adalimumab) injection, for subcutaneous use. September 2015.
- 5. EMA . Summary of product characteristics – Humira pre‐filled pen, pre‐filled syringe and vial. November 2015.
- 6. Schiff MH, Burmester GR, Kent JD, Pangan AL, Kupper H, Fitzpatrick SB, et al. Safety analyses of adalimumab (HUMIRA) in global clinical trials and US postmarketing surveillance of patients with rheumatoid arthritis. Ann Rheum Dis 2006; 65: 889–894. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Beck A, Reichert JM. Approval of the first biosimilar antibodies in Europe: a major landmark for the biopharmaceutical industry. MAbs 2013; 5: 621–623. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. EMA . Guideline on similar biological medicinal products. CHMP/437/04 Rev 1 ed2014.
- 9. EMA . Guideline on similar biological medicinal products containing biotechnology‐derived proteins as active substance: non‐clinical and clinical issues. EMEA/CHMP/BMWP/42832/2005 Rev1 ed2014.
- 10. FDA . Guidance for industry: scientific considerations in demonstrating Biosimilarity to a reference product (April 2015).
- 11. de Mora F. Biosimilar: what it is not. Br J Clin Pharmacol 2015; 80: 949–956. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Baji P, Péntek M, Czirják L, Szekanecz Z, Nagy G, Gulácsi L, et al. Efficacy and safety of infliximab‐biosimilar compared to other biological drugs in rheumatoid arthritis: a mixed treatment comparison. Eur J Health Econ 2014; 15 (Suppl 1): S53–S64. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Poddubnyy D, Rudwaleit M. Efficacy and safety of adalimumab treatment in patients with rheumatoid arthritis, ankylosing spondylitis and psoriatic arthritis. Expert Opin Drug Saf 2011; 10: 655–673. [DOI] [PubMed] [Google Scholar]
- 14. Poddubnyy D, Rudwaleit M. Adalimumab for the treatment of ankylosing spondylitis and nonradiographic axial spondyloarthritis ‐ a five‐year update. Expert Opin Biol Ther 2013; 13: 1599–1611. [DOI] [PubMed] [Google Scholar]
- 15. Tranter E, Peters G, Boyce M, Warrington S. Giving monoclonal antibodies to healthy volunteers in phase 1 trials: is it safe? Br J Clin Pharmacol 2013; 76: 164–172. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. FDA . Guidance for industry: skin irritation and sensitization testing of generic transdermal drug products (December 1999).
- 17. Guyatt GH, Townsend M, Keller JL, Singer J. Should study subjects see their previous responses: data from a randomized control trial. J Clin Epidemiol 1989; 42: 913–920. [DOI] [PubMed] [Google Scholar]
- 18. Center for drug evaluation and research for biologics evaluation and research. Clinical pharmacology and biopharmaceutics reviews. Adalimumab (40 mg subcutaneous administration). Available at http://www.fda.gov/downloads/Drugs/DevelopmentApprovalProcess/HowDrugsareDevelopedandApproved/ApprovalApplications/TherapeuticBiologicApplications/ucm092770.pdf (last accessed Dec 2016).
- 19. Bain BJ. Ethnic and sex differences in the total and differential white cell count and platelet count. J Clin Pathol 1996; 49: 664–666. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. EMA: guideline on the investigation of bioequivalence. CPMP/EWP/QWP/1401/98 Rev. 1 ed2010.
- 21. FDA: clinical pharmacology data to support a demonstration of Biosimilarity to a reference product (May 2014).
- 22. Tebbey PW, Varga A, Naill M, Clewell J, Venema J. Consistency of quality attributes for the glycosylated monoclonal antibody Humira® (adalimumab). MAbs 2015; 7: 805–811. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Wynne C, Altendorfer M, Sonderegger I, Gheyle L, Ellis‐Pegler R, Buschke S, et al. Bioequivalence, safety and immunogenicity of BI 695501, an adalimumab biosimilar candidate, compared with the reference biologic in a randomized, double‐blind, active comparator phase I clinical study (VOLTAIRE®‐PK) in healthy subjects. Expert Opin Investig Drugs 2016; 25: 1361–1370. [DOI] [PubMed] [Google Scholar]
- 24. Kaur P, Chow V, Zhang N, Moxness M, Kaliyaperumal A, Markus R. A randomised, single‐blind, single‐dose, three‐arm, parallel‐group study in healthy subjects to demonstrate pharmacokinetic equivalence of ABP 501 and adalimumab. Ann Rheum Dis 2016; pii: annrheumdis‐2015‐208914. doi:10.1136/annrheumdis‐2015‐208914 [Epub ahead of print]. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Jani RH, Gupta R, Bhatia G, Rathi G, Ashok Kumar P, Sharma R, et al. A prospective, randomized, double‐blind, multicentre, parallel‐group, active controlled study to compare efficacy and safety of biosimilar adalimumab (Exemptia; ZRC‐3197) and adalimumab (Humira) in patients with rheumatoid arthritis. Int J Rheum Dis 2015. doi:10.1111/1756‐185X.12711 [Epub ahead of print]. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Hyland E, Mant T, Vlachos P, Attkins N, Ullmann M, Roy S, et al. Comparison of the pharmacokinetics, safety, and immunogenicity of MSB11022, a biosimilar of adalimumab, with Humira® in healthy subjects. Br J Clin Pharmacol 2016; 82: 983–993. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. FDA news release: FDA approves Amjevita, a biosimilar to Humira. Available at http://www.fda.gov/NewsEvents/Newsroom/PressAnnouncements/ucm522243.htm (last accessed Dec 2016).
- 28. EMA . Summary of product characteristics – Humira pre‐filled syringe and pen. November 15 Sep 2016.
- 29. Finco D, Baltrukonis D, Clements‐Egan A, Delaria K, Gunn GR 3rd, Lowe J, et al. Comparison of competitive ligand‐binding assay and bioassay formats for the measurement of neutralizing antibodies to protein therapeutics. J Pharm Biomed Anal 2011; 54: 351–358. [DOI] [PubMed] [Google Scholar]
- 30. Bender N, Heilig CE, Dröll B, Wohlgemuth J, Armbruster FP, Heilig B. Immunogenicity, efficacy and adverse events of adalimumab in RA patients. Rheumatol Int 2007; 27: 269–274. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Figure S1 Scatterplots of pharmacokinetic parameters vs. anti‐drug antibody (ADA) activity presented by treatment group
Table S1 Serum concentrations of adalimumab after a 40‐mg dose for FKB327, EU‐Humira and US‐Humira
Table S2 Treatment‐emergent adverse events reported by <5% of subjects in any treatment group (system organ class and preferred term)
Table S3 Summary of ANOVA of primary and secondary adalimumab pharmacokinetic parameters: ratio of geometric least squares means with 90% confidence intervals (CI)