Abstract
Objectives
To explore associations between serum adalimumab level, treatment response and drug survival in order to identify optimal drug levels for therapeutic drug monitoring of adalimumab. Also, to assess the occurrence and risk factors of anti-drug antibody (ADAb) formation.
Methods
Non-trough adalimumab and ADAb levels were measured by automated fluorescence assays in serum collected after 3 months of adalimumab treatment in patients with RA, PsA or axial SpA (axSpA) included in the observational NOR-DMARD study. Treatment response was evaluated after 3 months and drug survival was evaluated during long-term follow-up.
Results
In 340 patients (97 RA, 69 PsA, 174 axSpA), the median adalimumab level was 7.3 mg/l (interquartile range 4.0–10.3). A total of 33 (10%) patients developed ADAbs. Findings were comparable across diagnoses. In RA and PsA, adalimumab levels ≥6.0 mg/l were associated with treatment response [odds ratio (OR) 2.2 (95% CI 1.0, 4.4)] and improved drug survival [hazard ratio 0.49 (95% CI 0.27, 0.80)]. In axSpA, a therapeutic level could not be identified, but higher adalimumab levels were associated with response. Factors associated with ADAb formation were previous bDMARD use, no methotrexate comedication and the use of adalimumab originator compared with GP2017.
Conclusion
Higher adalimumab levels were associated with a better response and improved drug survival for all diagnoses, with a suggested lower threshold of 6.0 mg/l for RA/PsA. This finding, the large variability in drug levels among patients receiving standard adalimumab dose and the high proportion of patients developing ADAbs encourages further investigations into the potential role of therapeutic drug monitoring of adalimumab.
Keywords: adalimumab, serum drug level, inflammatory joint disease, anti-drug antibodies, TNF inhibitors
Rheumatology key messages.
Higher adalimumab levels were associated with treatment response and improved drug survival across diagnoses.
The indicated lower threshold of adalimumab was 6.0 mg/l in RA/PsA.
ADAbs were found in 10% of patients, more commonly without methotrexate comedication, reducing treatment effect.
Introduction
TNF-α inhibitors (TNFis) and other biologic drugs have, together with novel treatment strategies such as treat to target, contributed to a revolution in the treatment of inflammatory joint diseases [1]. For patients with RA, PsA and axial SpA (axSpA), remission or inactive disease is now a realistic treatment goal. Nevertheless, despite current therapies and treatment strategies, a large proportion of patients do not respond sufficiently to therapy and approximately half of patients lose efficacy over time [2, 3]. Failure to maintain disease control has a major impact on quality of life and increases the risk of joint destruction for patients with peripheral arthritis. Loss of efficacy can be caused by underexposure to drugs, with or without development of anti-drug antibodies (ADAbs) [4–6]. Therapeutic drug monitoring (TDM), individualized dosing based on assessment of drug levels and ADAbs, is one strategy suggested to improve the effectiveness of TNFis [7]. TDM provides an opportunity to minimize both under- and overexposure to drugs. While underexposure can lead to loss of efficacy, overexposure increases costs and may predispose patients to adverse events [8]. In addition, timely identification of ADAbs enables early adjustment of treatment, possibly preventing a clinical flare [9]. In order to develop TDM algorithms, therapeutic ranges for the drug in question must be identified [10]. Recently published EULAR points to consider stated that one important barrier to using TDM in clinical care is the current lack of identified therapeutic ranges for most TNFis [10].
Adalimumab, a fully human monoclonal antibody, is the most commonly used TNFi worldwide in the treatment of several immune-mediated inflammatory diseases, including inflammatory joint diseases [11, 12]. While prior data suggest that adalimumab levels of 4–12 mg/l are associated with treatment response in RA and PsA patients [13–18], less is known regarding optimal serum adalimumab levels in axSpA [19–21].
Adalimumab has a high immunogenic potential, both in originator and biosimilar products, with ADAb development in 10–60% of patients, depending on the diagnosis and assay used for detection [6, 22]. Except for co-medication with methotrexate, which reduces the occurrence of ADAbs, little is known regarding factors associated with ADAb formation to adalimumab [23–25].
The main aim of this study was to explore associations between serum adalimumab levels, treatment response and drug survival in patients with inflammatory joint diseases, with the intention of identifying a therapeutic level. This would allow the development of TDM algorithms for adalimumab that can be validated in prospective clinical trials. Additionally, we aimed to explore the occurrence, risk factors and clinical implications of neutralizing ADAbs.
Methods
Study population
The Norwegian Antirheumatic Drug Registry (NOR-DMARD) (clinicaltrials.gov: NCT01581294) is a longitudinal multicentre observational study including adult patients with inflammatory joint diseases initiating therapy with biologic DMARDs (bDMARDs) [26, 27]. Clinical data are registered at baseline, 3, 6, 9 and 12 months and thereafter every 6 months. Biobank samples are collected at baseline and after 3 months.
In the current study we included patients with a clinical diagnosis of RA, PsA or axSpA with an available serum sample collected 3 months after initiating adalimumab. For assessment of treatment response, clinical data from baseline and 3 months were used. Information from the last registered follow-up visit was used in the assessment of drug survival.
All patients started adalimumab in a dose of 40 mg every second week. Patients were started on originator adalimumab up to 1 February 2020 and on the biosimilar GP2017 thereafter, based on a national annual tender system for biologic drugs in Norway [28].
The study was approved by an independent ethics committee (Regional Committees for Medical and Health Research Ethics South East; reference number 2011/1339) and all participants provided written informed consent.
Treatment response
In RA and PsA patients, disease activity was measured by the 28-joint Disease Activity Score with ESR (DAS28-ESR). Remission was defined as DAS28-ESR <2.6 and treatment response as EULAR good or moderate response [29, 30]. In patients with PsA, sensitivity analyses were performed using the 28-joint Disease Activity Score for Psoriatic Arthritis (DAPSA28), with DAPSA28 improvement ≥50% defined as treatment response [31, 32]. In axSpA patients, disease activity was measured by Ankylosing Spondylitis Disease Activity Score with CRP (ASDAS-CRP). Inactive disease was defined as ASDAS-CRP <1.3, treatment response as ASDAS major improvement (MI; change ≥2.0 units) or clinically important improvement (CII; change ≥1.1 units) [33]. ESR was used as a generic surrogate for disease activity.
Measurements of serum adalimumab levels and ADAbs
Non-trough serum samples stored at −80°C were analysed using a validated time-resolved fluorescence assay. The solid phase protein is human recombinant TNF and the tracer protein is europium-labelled protein A. Serum samples with adalimumab levels <3.0 mg/l were analysed with a drug-sensitive in-house fluorescence assay measuring neutralizing ADAbs, with human recombinant TNF as the solid phase protein and europium-labelled adalimumab F(ab′)2 as the tracer protein. ADAb levels ≥15 µg/l were defined as positive, with levels ≥50 µg/l considered moderate or high [9]. Both assays are fully automated on the AutoDELFIA immunoassay platform (PerkinElmer, Waltham, MA, USA).
Statistical analyses
Baseline characteristics were summarized with descriptive statistics. Comparisons of adalimumab levels and ADAb occurrence between treatment response and inactive disease/remission groups were analysed with the Mann–Whitney U test, χ2 test or independent samples t-test, as appropriate. Explorative concentration–effect analyses were used to suggest a possible therapeutic level, dividing the drug level range into segments with ≈21 and 22 patients in each (separately for RA/PsA and axSpA). Associations between the suggested therapeutic cut-offs and treatment response were further analysed by multivariable logistic regression, with response being the dependent variable and the independent variables being adalimumab cut-off, age, sex, prior use of a bDMARD (yes/no) and concomitant use of methotrexate. Sensitivity analyses using receiver operating characteristics (ROC) analyses were also performed to find the lower adalimumab cut-off by use of the Youden Index, maximizing the sum of sensitivity and specificity [34]. Drug survival was assessed using Kaplan–Meier curves and Cox proportional hazards regression analysis (adjusted for the same covariates as above). Patients who discontinued treatment due to remission, pregnancy or with missing information regarding the reason for drug discontinuation were censored at their last registered visit. Possible factors associated with adalimumab levels and ADAb formation were assessed with linear and logistic regression, respectively, with the independent variables being age, sex, prior use of a bDMARD (yes/no), concomitant methotrexate use, adalimumab type (originator or GP2017) and baseline disease activity. All tests were two-sided and performed at a 0.05 significance level.
The same outcomes were used in RA and PsA, and to increase the power they were handled as one disease group throughout the analyses. Subgroup analyses were performed.
Missing disease activity components were handled with median imputation and missing whole visits at 3 months were handled by next observation (6-month data) carried backwards. All analyses were done in Stata 16.1 (StataCorp, College Station, TX, USA) and GraphPad Prism 9.4.1 (GraphPad Software, San Diego, CA, USA).
Results
Study population and baseline characteristics
A total of 340 patients [97 RA, 69 PsA, 174 axSpA; mean age 46 years (s.d. 14); 181 (53%) female) starting adalimumab between 6 June 2012 and 27 October 2021 were eligible for inclusion in the present analyses (Supplementary Fig. S1, available at Rheumatology online). Originator adalimumab was used by 212 (62%) and GP2017 by 128 (38%) patients. Methotrexate co-medication was used by 121 patients (36%), mostly [108/121 (89%)] in patients with a diagnosis of RA or PsA (Table 1).
Table 1.
Baseline characteristics
| Characteristics | Total (N = 340) | RA (n = 97) | axSpA (n = 174) | PsA (n = 69) |
|---|---|---|---|---|
| Age, years, mean (SD) | 46 (14) | 53 (14) | 40 (12) | 51 (13) |
| Female, n (%) | 181 (53) | 74 (76) | 72 (41) | 35 (51) |
| Current smoker, n (%) | 28/232 (12) | 6/56 (11) | 17/135 (13) | 5/41 (12) |
| Disease duration, years, median (IQR) yearsa | 4.7 (1.0–11.9) | 5.5 (1.2–11.3) | 3.8 (0.8–12.6) | 4.4 (1.2–12.2) |
| Previous use of bDMARDs, n (%) | 112/333 (34) | 30/96 (31) | 61/170 (36) | 21/67 (31) |
| RF positivity, n (%) | 49/93 (53) | |||
| Anti-CCP positivity, n (%) | 65/96 (68) | |||
| HLA-B27 positivity, n (%) | 136/164 (83) | |||
| Peripheral involvement, n (%) | 52/173 (30) | |||
| Adalimumab, n (%) | ||||
| Originator | 212 (62) | 58 (60) | 113 (65) | 41 (59) |
| GP2017 | 128 (38) | 39 (40) | 61 (35) | 28 (41) |
| Co-medication, n (%) | ||||
| Methotrexateb | 121 (36) | 68 (70) | 13 (7) | 40 (59) |
| Sulfasalazine | 10 (3) | 5 (5) | 4 (2) | 1 (1) |
| Hydroxychloroquine | 2 (0.5) | 2 (2) | ||
| Leflunomide | 6 (2) | 4 (4) | 2 (3) | |
| Prednisolonec | 44 (14) | 34 (35) | 4 (2) | 6 (9) |
| Disease activity | ||||
| DAS28-ESR, mean (s.d.) | 3.7 (1.5) | 3.22 (1.3) | ||
| ASDAS-CRP, mean (s.d.) | 2.57 (1.0) | |||
| DAPSA28-CRP, mean (s.d.) | 17.1 (11.2) | |||
| Patient-reported global pain (VAS scale), median (IQR) | 44 (23–64) | 38 (16–58) | 45 (25–66) | 43 (26–60) |
VAS: visual analogue scale.
Available data in 180 patients.
Median dose of methotrexate per week: 20 mg (IQR 10–20).
Median prednisolone dose per day: 8.75 mg (IQR 5–15).
Adalimumab serum levels, treatment response and drug survival
The median adalimumab level at 3 months was 7.3 mg/l [interquartile range (IQR) 4.0–10.3], with comparable results across diagnoses (Fig. 1A). In RA/PsA patients, the adalimumab levels were not significantly different in patients with EULAR good or moderate response compared with non-responders [7.7 mg/l (IQR 5.0–10.3) vs 5.9 mg/l (IQR 1.9–10.2), P = 0.095] (Fig. 1B). In contrast, axSpA patients with ASDAS-CRP improvement (MI or CII) had higher adalimumab levels than patients without ASDAS-CRP improvement [7.9 mg/l (IQR 5.4–10.8) vs 6.6 mg/l (IQR 2.8–9.6), P = 0.014] (Fig 1B). Those in EULAR remission/ASDAS inactive disease had higher adalimumab levels than the remaining patients, both for RA/PsA [7.8 mg/l (IQR 5.1–10.5) vs 5.8 mg/l (IQR 1.2–9.0), P = 0.014] and axSpA [8.7 mg/l (IQR 6.1–10.9) vs 5.4 mg/l (IQR 2.2–8.6), P < 0.0001]) (Fig. 1C).
Figure 1.
Serum adalimumab levels by diagnosis, treatment response and remission/inactive disease at 3 months. Violin plot showing the probability density of the data at different values, smoothed by a kernel density estimator. Each data point is a participant, the solid orange line shows the group median and red dots are participants with ADAb formation. (A) Adalimumab levels for all participants and by diagnosis. (B and C) Adalimumab levels by (B) response to treatment (EULAR good and moderate response and ASDAS major and clinically important improvement) and (C) DAS28 remission/ASDAS inactive disease at 3 months. *Mann-Whitney U test
Based on explorative concentration–effect analyses in RA/PsA, the highest rates of response and remission were found in patients with adalimumab levels between 6.0 and 12.0 mg/l (Fig. 2A and 2B). ROC analyses supported this cut-off [6.0 mg/l (95% CI 1.3, 10.7)] with an area under the curve of 0.610 and a sensitivity of 68% and specificity of 55% (Supplementary Fig. S2A, available at Rheumatology online). Adalimumab levels ≥6.0 mg/l were associated with treatment response and improved drug survival, and this was consistent when adjusting for possible confounders [odds ratio (OR) for response 2.2 (95% CI 1.0, 4.4) (Supplementary Table S1, available at Rheumatology online) and hazard ratio (HR) 0.49 (95% CI 0.28, 0.85)] (Fig. 2C). Drug levels ≥12.0 mg/l were associated with a lower rate of response [OR 0.28 (95% CI 0.87, 0.93)] compared with levels between 6.0 and 12.0 mg/l. Separate graphs for RA and PsA and sensitivity analyses using DAPSA28 in PsA are shown in Supplementary Fig. S3, available at Rheumatology online.
Figure 2.
Treatment response, remission and drug survival in RA and PsA patients. (A and B) Drug-level range is divided into groups with ≈21 patients in each. Percent distribution of (A) EULAR response and (B) DAS28 remission in RA/PsA patients at 3 months according to adalimumab level. (C) Kaplan–Meier curve for 1.5 years drug survival stratified by adalimumab level at 3 months. Comparing RA/PsA patients with adalimumab ≥6 mg/l vs <6 mg/l, there was a significant difference in the survival estimates; P = 0.0003 (logrank)
Despite a concentration–effect relationship, we could not identify a therapeutic range for axSpA (Fig. 3A and 3B and Supplementary Fig. S2B, available at Rheumatology online), as the likelihood for both ASDAS response and inactive disease increased with increasing adalimumab [OR 1.2 (95% CI 1.02, 1.3) and OR 1.4 (95% CI 1.2, 1.7) per adalimumab level group increase, respectively]. The lowest response rate was seen in patients with adalimumab levels <1.5 mg/l [OR 0.2 (95% CI 0.08, 0.7) as compared with ≥1.5 mg/l]. Drug survival curves in axSpA patients for different drug-level groups are shown in Fig. 3C.
Figure 3.
Treatment response, inactive disease and drug survival in axSpA patients. (A and B) Drug-level range is divided into groups with ≈22 patients in each. Percent distribution of (A) ASDAS improvement and (B) ASDAS inactive disease in axSpA patients at 3 months according to adalimumab level. (C) Kaplan–Meier curve for 1.5 years drug survival stratified by adalimumab level at 3 months. There was no significant difference in the survival estimates; P = 0.082 (logrank)
Occurrence of ADAbs, treatment response and drug survival
At 3 months, 33 (10%) patients had developed ADAbs, [10 (10%) RA, 8 (12%) PsA and 15 (9%) axSpA patients], with a median ADAb level of 103 µg/l (IQR 60–182.5). Of patients with ADAb formation, 26/33 (79%) had moderate or high levels. The proportion with response to treatment was higher in patients without ADAb formation [178 (59%)] than in patients with ADAb formation [10 (31%)], with an OR of 2.3 (95% CI 1.04, 5.3). Further, patients with ADAb formation had a higher rate of drug discontinuation [HR 3.3 (95% CI 2.0, 5.3)] (Supplementary Fig S4, available at Rheumatology online).
Factors associated with adalimumab levels and ADAb formation
RA/PsA patients on concomitant treatment with methotrexate had significantly higher adalimumab levels compared with patients without co-medication [8.4 mg/l (IQR 5.5–11.0) vs 5.8 mg/l (IQR 1.2–8.7), P = 0.0002] and had less ADAb formation [5/108 (5%) vs 13/58 (22%), P < 0.001] (Table 2 and Supplementary Table S2, available at Rheumatology online). Further, patients without methotrexate co-medication had a higher drug discontinuation rate [HR 1.9 (95% CI 1.1, 3.3)].
Table 2.
Factors associated with ADAb formation
| Factors | Univariable OR (95% CI) | P-value | Multivariable OR (95% CI) | P-value |
|---|---|---|---|---|
| Age | 1.0 (0.98, 1.0) | 0.51 | 1.0 (0.98, 1.04) | 0.52 |
| Female sex | 1.8 (0.87, 4.0) | 0.11 | 2.1 (0.94, 4.6) | 0.071 |
| Previous use of one or more bDMARDs | 2.6 (1.3, 5.4) | 0.009 | 2.8 (1.3, 5.9) | 0.009 |
| Methotrexate co-medication | 0.37 (0.15, 0.93) | 0.034 | 0.31 (0.12, 0.82) | 0.018 |
| Adalimumab originator | 3.0 (1.2, 7.4) | 0.020 | 2.7 (1.1, 6.8) | 0.039 |
| ESR at baseline | 1.0 (0.98, 1.02) | 0.68 | 1.01 (0.98, 1.03) | 0.69 |
Significant results are in bold.
Patients treated with adalimumab originator had lower serum drug levels [6.4 mg/l (IQR 3.1–9.9) vs 8.3 mg/l (IQR 5.6–11.0), P = 0.0004] and a higher rate of ADAb formation [27/212 (13%) vs 6/128 (5%), P = 0.015] compared with patients treated with GP2017 (Fig. 4A). The between-group differences in serum drug levels and ADAb formation were consistent in multivariable regression analyses [β coefficient −1.45 (95% CI −2.4, −0.53) and OR 2.6 (95% CI 1.0, 6.7), with GP2017 as the reference group] and across diagnostic groups (Table 2 and Supplementary Table S2, available at Rheumatology online). There was no statistically significant difference in response to treatment or drug survival between adalimumab originator and GP2017 (Supplementary Table S3, available at Rheumatology online and Fig. 4B).
Figure 4.
Serum drug level and drug survival by originator and GP2017 adalimumab. (A) Violin plot showing the probability density of the data at different values, smoothed by a kernel density estimator. Each data point is a participant and the solid orange line shows the group median. Red dots are participants with ADAb formation. (B) Kaplan–Meier curve for 1.5 years drug survival stratified by originator and GP2017 adalimumab. There was no significant difference in the survival estimates; P = 0.16 (logrank)
Age, sex and baseline DASs were not associated with drug level or ADAb formation (Table 2 and Supplementary Table S2, available at Rheumatology online).
Discussion
This observational study is the first to explore associations between adalimumab serum levels and clinical outcomes across all adult inflammatory joint diseases. The current results suggest 6.0 mg/l as a lower therapeutic level for adalimumab in RA/PsA patients. These findings can contribute to the development of TDM algorithms for adalimumab that could be further tested in clinical trials.
In patients with axSpA, higher adalimumab levels were associated with response to treatment, but a specific therapeutic level could not be identified.
We found a large variability in adalimumab levels despite all patients receiving the same drug dose and a significant proportion of patients developing neutralizing ADAbs already at 3 months follow-up.
RA and PsA patients with serum drug levels ≥6.0 mg/l were more likely to respond to treatment at 3 months and had a lower risk of drug discontinuation. This finding corresponds to findings in previous studies where the suggested lower therapeutic level varied between 4 and 7 mg/l [13–18]. We suggest 6.0 mg/l as a lower therapeutic level for adalimumab, although we acknowledge that some individual patients will respond to therapy with lower levels depending on, for example, disease activity and disease phenotype. Based on our findings and prior data, 12.0 mg/l could be a possible upper limit for a therapeutic range [13, 15, 16, 18]. An upper limit of the therapeutic range is challenging to detect for TNFis, as it should take both risk of adverse events and drug costs into account. Despite these variables not being available in the present study, our results suggested 12.0 mg/l as a pragmatic upper limit for the therapeutic range, as adalimumab levels ≥12.0 mg/l were associated with lower response rates. However, as our assay measured free drug (as with most TNFi assays), high adalimumab levels in a patient not responsive to treatment may also indicate a different disease modality (where the pro-inflammatory effect of TNF is less important) not responsive to TNFi treatment and therefore with more free TNFi in the serum.
To our knowledge, this is the largest study to date examining the optimal adalimumab level in axSpA patients. Previous studies have diverging results [19, 20, 21, 35]. We could not identify a therapeutic range for axSpA, as the response rates increased with increasing adalimumab levels. However, Kaplan–Meier curves indicated that patients with adalimumab <1.5 mg/l had poorer drug survival. The lowest response rates were also found in this group, suggesting that patients should have levels at least >1.5 mg/l. Further, in patients with adalimumab >11.5 mg/l we saw a trend of more ASDAS major improvement, suggesting that some patients may benefit from very high levels. The outcome measures used in this study and other clinical trials within rheumatology are largely subjective, and factors like fibromyalgia could impact the measurements, making data on dose and response in these patients difficult to interpret. Also, the clinical axSpA diagnoses encompassed patients both with and without peripheral joint involvement, making for a somewhat heterogeneous population. These issues may account for the difficulty in determining therapeutic intervals in axSpA. The optimal therapeutic range may vary between patients due to individual differences in disease phenotype and disease activity. Future possibilities such as objective disease activity measurements, pharmacokinetic modelling and dashboard systems may enable individualized TDM [36].
In this study, 10% of patients had detectable neutralizing ADAbs after 3 months of adalimumab treatment, using a drug-sensitive inhibition assay measuring clinically relevant neutralizing ADAbs [6]. Most of them had levels >50 µg/l, considered clinically relevant and rarely transient [9]. Previous studies have suggested an occurrence of ADAb formation of 10–60%, dependent of the study population and assay [6, 10, 24, 37, 38]. In line with previous studies, patients with ADAb formation were less likely to respond to treatment in addition to having a higher risk of drug discontinuation [24, 38]. We found a comparable occurrence of ADAbs in axSpA as in RA/PsA patients. This is in contrast to the Norwegian Drug Monitoring Study A (NOR-DRUM-A; NCT03074656), a randomised controlled trial that tested the effectiveness of TDM when initiating infliximab treatment, where the risk of ADAb formation was lower in axSpA than RA/PsA patients [39]. One reason for this difference could be that the infliximab dose is lower in RA patients than in axSpA, while for adalimumab the dosage is the same for all diagnoses. In the NOR-DRUM-A trial, underexposure to drug over time or drug holidays were found to be a risk factor for subsequent ADAb development [39].
Patients with RA and PsA on concomitant treatment with methotrexate had higher adalimumab levels and a lower occurrence of ADAb formation, in addition to a lower rate of drug discontinuation. Methotrexate has demonstrated a favourable effect on the pharmacokinetics of adalimumab in previous studies [6, 13, 25, 40, 41] and is recommended to optimize the treatment effect of adalimumab in RA. The number of axSpA patients on methotrexate co-medication was too low in our cohort to assess any impact on ADAb formation and drug survival in this disease group.
Patients treated with adalimumab originator had lower serum drug levels and more frequent ADAb formation compared with GP2017. These results could not be explained by differences in baseline characteristics. Importantly, when assessing treatment effect and drug survival, there were no differences between adalimumab originator and GP2017. On seeking approval, a biosimilar product must submit immunogenicity assessments equivalent to the originator, with some difference in immunogenicity deemed acceptable [22]. Our findings are in line with the P17-301 trial, preceding the European Medicines Agency (EMA) approval of biosimilar GP2017, reporting slightly higher serum drug levels in GP2017-treated psoriasis patients through the whole study period [42]. Further, a difference in non-neutralizing ADAbs (45.1% in adalimumab originator, 35.8% in GP2017) was found, but when assessing neutralizing ADAbs only, the differences vanished. Safety and efficacy were equivalent and the differences in drug levels and ADAb frequency were regarded as clinically irrelevant [22, 42, 43]. A wide range of factors may affect detection of ADAb (assay design, time of sampling, population tested). Hence, pre-approval testing may not detect clinically meaningful differences in immunogenicity. The EMA, World Health Organization and US Food and Drug Administration guidelines recommend consideration of immunogenicity in pharmacovigilance for biologic agents. Here we report ADAb rates in a clinically meaningful context and perform subgroup analyses in three different arthritis groups. Our independent analyses revealed that while there was a significant difference in ADAb formation between the two versions of adalimumab, we did not observe any significant differences in their impact on clinical outcomes. These findings could alleviate some of the apprehensions surrounding the distinct immunogenicity of originator and biosimilar adalimumab.
The strengths of this study are the inclusion of three diagnoses, with a large group of axSpA patients, and real-life data collected 3 months after initiating adalimumab treatment, making it relevant for regular clinical care.
This study has some limitations. First, to ensure adequate statistical power, patients with RA and PsA were combined in the analysis. While we acknowledge that DAS28-ESR might not the optimal outcome measure for PsA, the conducted sensitivity analyses using the modified DAPSA, DAPSA28, yielded comparable results. Second, information on body weight or BMI, as well as adverse events, was lacking in this cohort and this information is relevant to serum drug levels. Obesity, with a BMI >30 kg/m2, has been shown to be associated with lower serum drug levels of adalimumab and etanercept and may therefore contribute to the observed variations in serum drug levels [40]. Third, the serum drug levels are non-trough, and the timing of the last injection was not registered. However, drug levels are reasonably stable through an injection cycle for adalimumab and other TNFis administered subcutaneously [44, 45]. This is reassuring as non-trough sampling is feasible in a clinical setting, while trough levels for subcutaneous drugs are difficult to obtain [40]. Finally, due to a lack of data, we were unable to explore the associations between smoking, drug holidays and adherence on serum drug levels and ADAb formation.
Conclusions
Higher adalimumab levels were associated with a better response and improved drug survival for all diagnoses, with 6.0 mg/l suggested as a lower therapeutic threshold in RA and PsA. As early as 3 months, 10% of patients had developed neutralizing ADAbs and these were associated with a lack of response to treatment and reduced drug survival. The considerable variability in drug levels in patients on the same standard dose, the correlation between these and treatment effectiveness and the high proportion of patients developing ADAbs all underscore the need for additional research to investigate the potential role of TDM in adalimumab treatment regimens. The present study will contribute to the development of TDM algorithms for adalimumab in patients with inflammatory joint diseases.
Supplementary Material
Acknowledgements
We would like to thank the patients participating in the study and are very grateful for the time and effort they have invested in the project. We would also like to thank personnel at the Department of Medical Biochemistry, Oslo University Hospital, Radiumhospitalet, study personnel involved at the Division of Rheumatology and Research at Diakonhjemmet Hospital and at Lillehammer Hospital for Rheumatic Diseases, especially Eva Melbø.
Contributor Information
Ingrid Jyssum, Center for Treatment of Rheumatic and Musculoskeletal Diseases (REMEDY), Diakonhjemmet Hospital, Oslo, Norway; Institute of Clinical Medicine, Faculty of Medicine, University of Oslo, Oslo, Norway.
Johanna E Gehin, Department of Medical Biochemistry, Oslo University Hospital, Oslo, Norway.
Joseph Sexton, Center for Treatment of Rheumatic and Musculoskeletal Diseases (REMEDY), Diakonhjemmet Hospital, Oslo, Norway.
Eirik Klami Kristianslund, Center for Treatment of Rheumatic and Musculoskeletal Diseases (REMEDY), Diakonhjemmet Hospital, Oslo, Norway.
Yi Hu, Lillehammer Hospital for Rheumatic Diseases, Lillehammer, Norway.
David John Warren, Department of Medical Biochemistry, Oslo University Hospital, Oslo, Norway.
Tore K Kvien, Center for Treatment of Rheumatic and Musculoskeletal Diseases (REMEDY), Diakonhjemmet Hospital, Oslo, Norway; Institute of Clinical Medicine, Faculty of Medicine, University of Oslo, Oslo, Norway.
Espen A Haavardsholm, Center for Treatment of Rheumatic and Musculoskeletal Diseases (REMEDY), Diakonhjemmet Hospital, Oslo, Norway; Institute of Clinical Medicine, Faculty of Medicine, University of Oslo, Oslo, Norway.
Silje Watterdal Syversen, Center for Treatment of Rheumatic and Musculoskeletal Diseases (REMEDY), Diakonhjemmet Hospital, Oslo, Norway.
Nils Bolstad, Department of Medical Biochemistry, Oslo University Hospital, Oslo, Norway.
Guro Løvik Goll, Center for Treatment of Rheumatic and Musculoskeletal Diseases (REMEDY), Diakonhjemmet Hospital, Oslo, Norway.
Supplementary material
Supplementary material is available at Rheumatology online.
Data availability
A de-identified patient data set can be made available to researchers upon reasonable request. The data will only be made available after submission of a project plan outlining the reason for the request and any proposed analyses and will have to be approved by the NOR-DMARD steering group. Project proposals can be submitted to the corresponding author. Data sharing will have to follow appropriate regulations.
Authors’ contributions
All authors critically revised the report and approved the final submitted version and take responsibility for the completeness and accuracy of the data and analyses. All authors had full access to all the data in the study and made the final decision to submit the manuscript for publication. I.J., G.L.G., S.W.S., N.B., J.E.G. and J.S. verified the underlying data, interpreted the data and drafted the report. I.J., G.L.G., S.W.S., N.B. and J.E.G. conceived and designed the study. D.J.W. developed all assay reagents and established the drug-level assay. I.J., N.B. and J.E.G. performed the drug-level and ADAb analyses. E.A.H., T.K.K. and E.K.K. contributed to the study conception and design. Y.H. contributed to data collection.
Funding
This work was supported by the South-Eastern Norway Regional Health Authority (project number 2020017), the Research Council of Norway (project number 328657) and the Olav Thon Foundation. The NOR-DMARD registry has been financially supported by pharmaceutical companies. The funding sources were not involved in the study design; collection, analysis and interpretation of data; writing of the report; or the decision to submit the paper for publication.
Disclosure statement: T.K.K. has received grants from AbbVie, Amgen, Bristol-Myers Squibb, Galapagos, Novartis, Pfizer and UCB; consulting fees from AbbVie, Amgen, Celltrion, Gilead, Novartis, Pfizer, Sandoz and UCB; and participated on speaker bureaus for Grünenthal, UCB and Sandoz. G.L.G. has received speaker fees from AbbVie/Abbott, Galapagos, Pfizer and UCB and participated on advisory boards for Pfizer, AbbVie/Abbott, Galapagos, Pfizer and UCB. E.A.H. has received speaker/consultant fees from Pfizer, AbbVie, Gilead, UCB Pharma, Galapagos, Eli Lilly, Novartis and Boehringer Ingelheim. Y.H. has participated on speaker bureaus for Boehringer. I.J., S.W.S., J.E.G., E.K.K., N.B., J.S. and D.J.W. had nothing to disclose.
References
- 1. Smolen JS, Landewe RBM, Bergstra SA et al. EULAR recommendations for the management of rheumatoid arthritis with synthetic and biological disease-modifying antirheumatic drugs: 2022 update. Ann Rheum Dis 2023;82:3–18. [DOI] [PubMed] [Google Scholar]
- 2. Arora A, Mahajan A, Spurden D, Boyd H, Porter D. Long-term drug survival of TNF inhibitor therapy in RA patients: a systematic review of european national drug registers. Int J Rheumatol 2013;2013:764518. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Movahedi M, Hepworth E, Mirza R et al. Discontinuation of biologic therapy due to lack/loss of response and adverse events is similar between TNFi and non-TNFi class: results from a real-world rheumatoid arthritis cohort. Semin Arthritis Rheum 2020;50:915–22. [DOI] [PubMed] [Google Scholar]
- 4. Bartelds GM, Krieckaert CL, Nurmohamed MT et al. Development of antidrug antibodies against adalimumab and association with disease activity and treatment failure during long-term follow-up. JAMA 2011;305:1460–8. [DOI] [PubMed] [Google Scholar]
- 5. Mehta P, Manson JJ. What is the clinical relevance of TNF inhibitor immunogenicity in the management of patients with rheumatoid arthritis? Front Immunol 2020;11:589. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Gehin JE, Goll GL, Brun MK et al. Assessing immunogenicity of biologic drugs in inflammatory joint diseases: progress towards personalized medicine. BioDrugs 2022;36:731–48. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Krieckaert C, Hernández-Breijo B, Gehin JE et al. Therapeutic drug monitoring of biopharmaceuticals in inflammatory rheumatic and musculoskeletal disease: a systematic literature review informing EULAR points to consider. RMD Open 2022;8:e002216. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Krieckaert CL, Nair SC, Nurmohamed MT et al. Personalised treatment using serum drug levels of adalimumab in patients with rheumatoid arthritis: an evaluation of costs and effects. Ann Rheum Dis 2015;74:361–8. [DOI] [PubMed] [Google Scholar]
- 9. Syversen SW, Jørgensen KK, Goll GL et al. Effect of therapeutic drug monitoring vs standard therapy during maintenance infliximab therapy on disease control in patients with immune-mediated inflammatory diseases: a randomized clinical trial. JAMA 2021;326:2375–84. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Krieckaert CL, van Tubergen A, Gehin JE et al. EULAR points to consider for therapeutic drug monitoring of biopharmaceuticals in inflammatory rheumatic and musculoskeletal diseases. Ann Rheum Dis 2023;82:65–73. [DOI] [PubMed] [Google Scholar]
- 11. Urquhart L. Top companies and drugs by sales in 2021. Nat Rev Drug Discov 2022;21:251. [DOI] [PubMed] [Google Scholar]
- 12. Mease PJ. Adalimumab in the treatment of arthritis. Ther Clin Risk Manag 2007;3:133–48. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Pouw MF, Krieckaert CL, Nurmohamed MT et al. Key findings towards optimising adalimumab treatment: the concentration-effect curve. Ann Rheum Dis 2015;74:513–8. [DOI] [PubMed] [Google Scholar]
- 14. Hum RM, Ho P, Nair N et al. Non-Trough adalimumab and certolizumab drug levels associated with a therapeutic EULAR response in adherent patients with rheumatoid arthritis. Rheumatology (Oxford) 2023;62:2090–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Jani M, Chinoy H, Barton A, for OUTPASS. Association of pharmacological biomarkers with treatment response and longterm disability in patients with psoriatic arthritis: results from OUTPASS. J Rheumatol 2020;47:1204–8. [DOI] [PubMed] [Google Scholar]
- 16. Vogelzang EH, Kneepkens EL, Nurmohamed MT et al. Anti-adalimumab antibodies and adalimumab concentrations in psoriatic arthritis; an association with disease activity at 28 and 52 weeks of follow-up. Ann Rheum Dis 2014;73:2178–82. [DOI] [PubMed] [Google Scholar]
- 17. Ducourau E, Ternant D, Lequerré T et al. Towards an individualised target concentration of adalimumab in rheumatoid arthritis. Ann Rheum Dis 2014;73:1428–9. [DOI] [PubMed] [Google Scholar]
- 18. Rosas J, Llinares-Tello F, de la Torre I et al. Clinical relevance of monitoring serum levels of adalimumab in patients with rheumatoid arthritis in daily practice. Clin Exp Rheumatol 2014;32:942–8. [PubMed] [Google Scholar]
- 19. Marsman AF, Kneepkens EL, Ruwaard J et al. Search for a concentration–effect curve of adalimumab in ankylosing spondylitis patients. Scand J Rheumatol 2016;45:331–4. [DOI] [PubMed] [Google Scholar]
- 20. Ding X, Zhu R, Wu J et al. Early adalimumab and anti-adalimumab antibody levels for prediction of primary nonresponse in ankylosing spondylitis patients. Clin Transl Sci 2020;13:547–54. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Paramarta JE, Baeten DL. Adalimumab serum levels and antidrug antibodies towards adalimumab in peripheral spondyloarthritis: no association with clinical response to treatment or with disease relapse upon treatment discontinuation. Arthritis Res Ther 2014;16:R160. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Kurki P, Barry S, Bourges I, Tsantili P, Wolff-Holz E. Safety, immunogenicity and interchangeability of biosimilar monoclonal antibodies and fusion proteins: a regulatory perspective. Drugs 2021;81:1881–96. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Burmester GR, Kivitz AJ, Kupper H et al. Efficacy and safety of ascending methotrexate dose in combination with adalimumab: the randomised CONCERTO trial. Ann Rheum Dis 2015;74:1037–44. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Thomas SS, Borazan N, Barroso N et al. Comparative immunogenicity of TNF inhibitors: impact on clinical efficacy and tolerability in the management of autoimmune diseases. A systematic review and meta-analysis. BioDrugs 2015;29:241–58. [DOI] [PubMed] [Google Scholar]
- 25. Goss SL, Klein CE, Jin Z et al. Methotrexate dose in patients with early rheumatoid arthritis impacts methotrexate polyglutamate pharmacokinetics, adalimumab pharmacokinetics, and efficacy: pharmacokinetic and exposure-response analysis of the CONCERTO trial. Clin Ther 2018;40:309–19. [DOI] [PubMed] [Google Scholar]
- 26. Kvien TK, Heiberg Lie E, Kaufmann C et al. A Norwegian DMARD register: prescriptions of DMARDs and biological agents to patients with inflammatory rheumatic diseases. Clin Exp Rheumatol 2005;23:S188–94. [PubMed] [Google Scholar]
- 27. Olsen IC, Haavardsholm EA, Moholt E, Kvien TK, Lie E. NOR-DMARD data management: implementation of data capture from electronic health records. Clin Exp Rheumatol 2014;32:S-158–62. [PubMed] [Google Scholar]
- 28. Goll GL, Kvien TK. An opportunity missed: biosimilars in the United States. Arthritis Rheumatol 2020;72:1046–8. [DOI] [PubMed] [Google Scholar]
- 29. Prevoo ML, van 't Hof MA, Kuper HH et al. Modified disease activity scores that include twenty-eight-joint counts. Development and validation in a prospective longitudinal study of patients with rheumatoid arthritis. Arthritis Rheum 1995;38:44–8. [DOI] [PubMed] [Google Scholar]
- 30. van Gestel AM, Haagsma CJ, van Riel PL. Validation of rheumatoid arthritis improvement criteria that include simplified joint counts. Arthritis Rheum 1998;41:1845–50. [DOI] [PubMed] [Google Scholar]
- 31. Michelsen B, Sexton J, Smolen JS et al. Can disease activity in patients with psoriatic arthritis be adequately assessed by a modified Disease Activity index for PSoriatic Arthritis (DAPSA) based on 28 joints? Ann Rheum Dis 2018;77:1736–41. [DOI] [PubMed] [Google Scholar]
- 32. Schoels MM, Aletaha D, Alasti F, Smolen JS. Disease activity in psoriatic arthritis (PsA): defining remission and treatment success using the DAPSA score. Ann Rheum Dis 2016;75:811–8. [DOI] [PubMed] [Google Scholar]
- 33. Machado P, Landewé R, Lie E et al. ; Assessment of SpondyloArthritis international Society. Ankylosing Spondylitis Disease Activity Score (ASDAS): defining cut-off values for disease activity states and improvement scores. Ann Rheum Dis 2011;70:47–53. [DOI] [PubMed] [Google Scholar]
- 34. Youden WJ. Index for rating diagnostic tests. Cancer 1950;3:32–5. [DOI] [PubMed] [Google Scholar]
- 35. Senabre Gallego JM, Rosas J, Marco-Mingot M et al. ; AIRE-MB Group. Clinical relevance of monitoring serum adalimumab levels in axial spondyloarthritis. Rheumatol Int 2019;39:841–9. [DOI] [PubMed] [Google Scholar]
- 36. Irving PM, Gecse KB. Optimizing therapies using therapeutic drug monitoring: current strategies and future perspectives. Gastroenterology 2022;162:1512–24. [DOI] [PubMed] [Google Scholar]
- 37. Borrega R, Araújo C, Aguiam N et al. Systematic review and principal components analysis of the immunogenicity of adalimumab. BioDrugs 2021;35:35–45. [DOI] [PubMed] [Google Scholar]
- 38. Strand V, Balsa A, Al-Saleh J et al. Immunogenicity of biologics in chronic inflammatory diseases: a systematic review. BioDrugs 2017;31:299–316. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39. Brun MK, Goll GL, Jørgensen KK et al. Risk factors for anti-drug antibody formation to infliximab: secondary analyses of a randomised controlled trial. J Intern Med 2022;292:477–91. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40. Jani M, Chinoy H, Warren RB et al. ; Biologics in Rheumatoid Arthritis Genetics and Genomics Study Syndicate Collaborators. Clinical utility of random anti-tumor necrosis factor drug-level testing and measurement of antidrug antibodies on the long-term treatment response in rheumatoid arthritis. Arthritis Rheumatol 2015;67:2011–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41. Krieckaert CL, Nurmohamed MT, Wolbink GJ. Methotrexate reduces immunogenicity in adalimumab treated rheumatoid arthritis patients in a dose dependent manner. Ann Rheum Dis 2012;71:1914–5. [DOI] [PubMed] [Google Scholar]
- 42. Blauvelt A, Lacour JP, Fowler JF Jr et al. Phase III randomized study of the proposed adalimumab biosimilar GP2017 in psoriasis: impact of multiple switches. Br J Dermatol 2018;179:623–31. [DOI] [PubMed] [Google Scholar]
- 43. Assessment report Hyrimoz. European Medicines Agency (EMA). 2018. https://www.ema.europa.eu/en/documents/assessment-report/hyrimoz-epar-public-assessment-report_en.pdf (2 February 2023, date last accessed).
- 44. Ungar B, Engel T, Yablecovitch D et al. Prospective observational evaluation of time-dependency of adalimumab immunogenicity and drug concentrations: the POETIC study. Am J Gastroenterol 2018;113:890–8. [DOI] [PubMed] [Google Scholar]
- 45. Schreiber S, Ben-Horin S, Leszczyszyn J et al. Randomized controlled trial: subcutaneous vs intravenous infliximab CT-P13 maintenance in inflammatory bowel disease. Gastroenterology 2021;160:2340–53. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
A de-identified patient data set can be made available to researchers upon reasonable request. The data will only be made available after submission of a project plan outlining the reason for the request and any proposed analyses and will have to be approved by the NOR-DMARD steering group. Project proposals can be submitted to the corresponding author. Data sharing will have to follow appropriate regulations.