REVIEW ARTICLE
Pharmacokinetics of Adalimumab in Inflammatory Bowel Diseases: A Systematic Review and Meta-analysis Stephane Paul, PhD,* Amelie Carla Moreau, PhD,* Emilie Del Tedesco, MD,† Melanie Rinaudo, MD, PhD,* Jean-Marc Phelip, MD, PhD,† Christian Genin, MD, PhD,* Laurent Peyrin-Biroulet, MD, PhD,‡ and Xavier Roblin, MD, PhD†
Background: The aim of this meta-analysis was to explore the magnitude of the association between pharmacokinetics of adalimumab and clinical response in patients with inflammatory bowel disease. Methods: A literature search was performed up to December 2013. MEDLINE, EMBASE, Cochrane, and meeting abstracts were reviewed. Studies were included if they analyzed the association of trough levels of adalimumab (TRA) or antibodies against adalimumab (AAA) with clinical response in adult or pediatric inflammatory bowel disease. A Mantel–Haenszel pooled risk estimate provided a measure of the association.
Results: Fourteen studies enrolling 1941 patients with inflammatory bowel disease were included in the systematic review. Thirteen studies analyzed clinical outcomes according to TRA. In only 1 study, there was no correlation between high TRA and clinical response. Six of the 7 studies reported a negative correlation between AAA and clinical outcomes. Six studies enrolling 536 patients (Crohn’s disease [CD] only) met the meta-analysis inclusion criteria. The pooled odds ratio (OR) for loss of clinical response to adalimumab in patients with CD (N ¼ 4) with positive AAAs was 10.15 (95% confidence interval [CI]: 3.90–26.40, P , 0.0001). Patients with CD with TRA over a predefined cutoff were more likely to be in clinical remission with an OR of 2.6 (95% CI: 1.79–3.77, P , 0.0001). The association was stronger if the analysis was limited to the adult population (N ¼ 3, OR: 7.05, 95% CI: 3.58–13.9, P , 0.0001). Conclusions: The presence of AAA is associated with a higher risk of loss of clinical response to adalimumab, whereas high TRA is associated with greater clinical response rates in CD. More data are needed in ulcerative colitis. (Inflamm Bowel Dis 2014;20:1288–1295) Key Words: inflammatory bowel diseases, adalimumab trough levels, antibodies to adalimumab, therapeutic drug monitoring, clinical response
T
hree anti-tumor necrosis factor (TNF) agents, namely, infliximab, adalimumab, and certolizumab pegol, are approved by the Food and Drug Administration (FDA) for the treatment of refractory luminal Crohn’s disease (CD).1 Infliximab, adalimumab, and golimumab are approved by FDA for the treatment of refractory ulcerative colitis (UC).2–5 Measuring antibodies against anti-TNF therapy and drug trough levels are increasingly used to improve disease outcomes in inflammatory bowel diseases (IBD).6,7 In a meta-analysis including 13 studies and 1378 patients with IBD,8 the presence of antibodies against infliximab (ATIs) was associated with a significantly higher risk of loss of clinical response to infliximab {relative risk of loss of Received for publication February 1, 2014; Accepted February 27, 2014. From the *Laboratoire d’Immunologie et d’Immunomonitoring, CHU de SaintEtienne, Saint-Etienne, France; †Service de Gastrologie-Entérologie-Hépatologie, CHU de Saint-Etienne, Saint-Etienne, France; and ‡Department of Gastroenterology, Inserm U954, University Hospital of Nancy, Université de Lorraine, Nancy, France. L. Peyrin-Biroulet received Lecture and/or consulting fees from MSD, Abbvie, UCB-pharma, and X. Roblin received Lecture and consulting fees from Merck, Abbvie, and Theradiag. The remaining authors have no conflicts of interest to disclose. Reprints: Xavier Roblin, MD, PhD, Department of Gastroenterology, CHU de Saint-Etienne, Avenue Albert Raymond, Saint-Etienne 42023, France (e-mail: xavier. [email protected]). Copyright © 2014 Crohn’s & Colitis Foundation of America, Inc. DOI 10.1097/MIB.0000000000000037 Published online 14 May 2014.
1288
| www.ibdjournal.org
response of 2.2 (95% confidence interval [CI]: 0.5–9.0, P ¼ 0.3)} and lower serum infliximab levels (trough levels of infliximab). In a second meta-analysis including 8 studies, the presence of ATIs was associated with a higher risk of acute infusion reactions, but not delayed hypersensitivity reactions, in patients with IBD.9 Concomitant immunomodulators use reduced this risk.9 Based on these findings, Ordas et al10 concluded that incorporation of therapeutic drug monitoring into clinical practice may allow clinicians to optimize treatment by maintaining effective drug concentrations over time. Adalimumab is a fully human anti-TNF. Recently, several studies investigated the association of pharmacokinetics of adalimumab with clinical benefits in IBD.11–18 However, some studies were of small sample size, and some of them provided conflicting results.11,12,19 Our aim was first to review available evidence regarding the association between pharmacokinetics of adalimumab and clinical response in IBD before conducting a meta-analysis of studies fulfilling inclusion criteria.
METHODS Study Selection We conducted a comprehensive computerized search for English and non-English language publications listed in the electronic databases of MEDLINE (source PUBMED, 1966 to Inflamm Bowel Dis Volume 20, Number 7, July 2014
Inflamm Bowel Dis Volume 20, Number 7, July 2014
December 2013), the Cochrane Library (Issue 12, 2013), and EMBASE (1980 to December 2013). We searched and combined the following terms: “drug monitoring, inflammatory bowel disease, Crohn’s disease, ulcerative colitis, adalimumab, drug level, antibodies, outcomes, treatment response, impact of drug level, impact of antibody formation.” We also hand-searched abstracts from the annual meetings of Digestive Disease Week and the United European Gastroenterology Week between 2005 and 2013, as well as references from review articles, meta-analysis studies, and published randomized controlled trials so as to identify additional articles. Authors of identified studies and pharmaceutical companies were also contacted to retrieve missing or unpublished data. Studies were selected based on the title and/or the abstract by 2 researchers (A.C.M. and X.R.). The reviewers checked whether inclusion and exclusion criteria were met and, in the event of uncertainty, the full text of articles was retrieved and reviewed. To be included in the systematic review, the focus of the study had to be the reporting of clinical response under adalimumab. We included all studies (controlled trials, observational studies, and cohort studies) that reported antibodies against adalimumab (AAA) and/or trough levels of adalimumab (TRA) and clinical outcomes in patients with IBD who were treated with adalimumab. Where AAA was studied, the primary outcome measure was the loss of response, defined as relapse of clinical symptoms in patients who were in clinical remission from, or had responded to, adalimumab. No prespecified score was used to determine this outcome, as we expected there would be no universal definition of the outcome across the studies. Where TRA were studied, the primary outcome measure was clinical response. Studies were excluded if they (1) were review articles, (2) the population presented was non-IBD patients, (3) did not measure either TRA or AAA, (4) did not report clinical outcomes, and (5) were studies using other anti-TNFs (infliximab or certolizumab). If there was any suspicion of a duplicate study, the most recent study with the largest cohort was considered for inclusion. There were additional inclusion criteria for the metaanalysis. Studies were included if they reported clinical outcomes according to AAA status or TRA with a predefined cutoff, thus allowing the construction of 2 · 2 tables.
Outcome Assessment Outcome measures were defined a priori. The efficacy end points were clinical remission, absence of flare-up, deep remission, or normalization of C-reactive protein (CRP). The definition of clinical activity varied across studies (Table 2).
Data Extraction Data abstraction was carried out independently by 2 investigators (A.C.M. and X.R.) using standardized data collection forms. If needed, authors were contacted regarding specific questions relating to their study. The independent reviewers conferred after data extraction was complete, discrepancies were identified, and review of the relevant article with a third author (S.P.) led to consensus.
Pharmacokinetics of Adalimumab
The following data were retrieved where possible: journal, year, author, number of patients in the study, method of selection of cohorts, drug regimen, assay type used to measure TRA and/or AAA, numbers of AAA-positive and AAA-negative patients, numbers of TRA over the cutoff and under the cutoff, methods of measurement of clinical outcomes, and reporting of outcomes used. Assessment of quality of randomized controlled trials and observational studies was performed using The GRADE (Grading of Recommendations Assessment, Development and Evaluation) criteria.20 For observational studies, methodological quality was assessed by determining the eligibility criteria, degree of measurement of both exposure and outcome, control for confounders, and completeness of follow-up.21
Data Synthesis and Statistical Analysis The primary goal of the meta-analysis was to assess the association between AAA or adalimumab trough levels and clinical remission in patients with IBD. Secondary objectives were to evaluate whether or not the association was influenced by differences in study design such as the presence of a selection and/ or information bias. A Mantel–Haenszel pooled risk estimate provided a measure of that association. Raw data from included studies (absolute numbers) were used to construct 2 · 2 contingency tables, and unadjusted pooled odd ratios (ORs) with 95% CIs were calculated using Review Manager (RevMan 5.1; Nordic Cochrane Centre, Copenhagen, Denmark) for dichotomous outcomes. Forest plots and funnels plots were generated using R version 3.0.2 (R, University of Auckland, Auckland, New Zealand). The Q test was used to assess for heterogeneity and I2 statistic to quantify the percentage of heterogeneity because of between-study variations; a value of P , 0.10 was considered statistically significant.22 Any heterogeneity identified prompted subgroup analysis in an attempt to explain these findings. If needed, a random effects model by DerSimonian and Laird23 was used to account for variations between studies and gives a more conservative pooled estimate in case where heterogeneity between studies could not be explained. A funnel plot was generated and checked for evidence of asymmetry, and therefore possible publication bias or other small study effects, using the Egger’s test, a P value ,0.10 was regarded as significant.24 Sensitivity analyses were performed for all outcomes where 3 or more studies were included. Sensitivity analysis included sensitivity based on individual studies, sample size, drug regimen, length of the study, outcome measure, and drugs used in the control arm. Significance levels were set at a P value of ,0.05.
RESULTS Systematic Review of the Association Between Pharmacokinetics of Adalimumab and Clinical Benefit Database searches identified 499 references, and 14 articles were included in the systematic review (Fig. 1): 11 studies in CD, www.ibdjournal.org |
1289
Inflamm Bowel Dis Volume 20, Number 7, July 2014
Paul et al
2 in UC,3,25 and 1 included both CD and UC patients12 (Table 1). Two studies included patients with pediatric CD.13,26 Twelve studies included adult patients with IBD.3,11,12,14–19,25,27,28 Six studies evaluated disease activity in patients on adalimumab treatment according to TRA and AAA,11,12,14–16,28 7 studies analyzed disease activity according to TRA with no reported data on AAA,3,13,17–19,25,26 and 1 study analyzed disease activity according to AAA without data on TRA.27 Four randomized controlled trials,3,18,19,26 4 prospective crosssectional studies,12,14–16 5 observational studies,11,13,17,25,28 and 1 retrospective study17 were included in the systematic review. A quiescent disease was defined by clinical remission in 10 studies11–16,18,19,25,28 by CRP normalization in 1 study,14 by deep remission defined by a clinical remission and CRP normalization17 in 1 study, and 1 study evaluated clinical response after adalimumab dose escalation.25 One study analyzed the correlation between clinical response and AAA without data on TRA in CD.27
was performed in the one by Sandborn et al3 Ferrante et al25 analyzed TRA after adalimumab dose escalation, and Roblin et al12 analyzed the association between TRA and clinical remission by pooling UC and CD. Ten CD11,13–19,26,28 studies looked at the association between TRA and clinical response. Only 1 study reported no increased rate of clinical response among CD patients with higher TRA.19
Identification of an Optimal Cutoff for TRA for Predicting Clinical Remission Area under curve was used to identify the optimal cutoff for TRA for predicting clinical remission in 7 studies.11,12,14–16,18,19 A cutoff value was identified in 4 studies, ranging from 4.8 to 5.9 mg/ mL.12,14–16 For Karmiris et al,11 patients who had TRA ,0.33 mg/mL at least once (sensitivity, 95%; positive predictive value, 81%) demonstrated significantly less sustained clinical benefit than patients never showing such low trough serum concentration (log-rank test; P , 0.01).
Pediatric Versus Adult Studies Two pediatric studies (N ¼ 326) included only CD patients.13,26 In both studies, there was an increased rate of remission and response among patients with TRA above the median concentration.13,26 No data about AAA were reported in these 2 studies.13,26 In the adult population (N ¼ 1615 IBD patients), there was a correlation between high TRA and clinical response in all 11 studies3,11,12,14–18,25,28 except the study by Chiu et al19 (Table 1).
Impact of Concomitant Immunosuppressive Treatment
Crohn’s Disease Versus Ulcerative Colitis
In 4 studies11,12,18,19 where the authors analyzed the impact of concomitant immunosuppressive therapy on disease outcome, it did not affect treatment efficacy and did not influence TRA.11,12,18,19 However, time to adalimumab dose escalation was longer under combination therapy (17 weeks versus 12 weeks; P ¼ 0.008).8 Chiu et al19 reported higher but not significantly different TRA at weeks 24 and 56 among patients on combination therapy.
Two studies included only UC patients3,25 and 1 combined UC and CD patients.12 All 3 studies found an association between TRA and clinical response, even though no statistical comparison
Impact of Antibodies Against Adalimumab on Disease Outcomes Nine studies investigated so far the percentage of positive AAA among adalimumab-treated IBD patients,11,12,14–16,18,19,27,28 7 studies11,12,14–16,27,28 have been published as full-length papers and all of them included only adult patients. The percentage of positive AAA varied between 0.04% and 35%.11,12,14–16,18,19,27,28 Differences between available assays and differing methodologies may explain these variations. Almost all studies reported a negative correlation between AAA levels and TRA.11,12,14–16,27,28 Six of the 7 studies found a negative correlation between AAA and clinical outcomes (5 with clinical disease activity11,12,15,16,27 and 2 with CRP levels).14,15 These results were not confirmed in 1 study.28 Concomitant immunosuppressive therapy did not affect treatment outcome and did not influence AAA formation.11,12,16,27
Meta-analysis of the Association Between Pharmacokinetics of Adalimumab and Clinical Benefit
FIGURE 1. Flow chart of study selection.
1290
| www.ibdjournal.org
Of the 14 studies selected and included in the systematic review, only 6 studies met the inclusion criteria and were included in the meta-analysis.12–14,16,26,27 Three studies evaluated both TRA and AAA,12,14,16 2 only 13,26 and 1 study only AAA.27 Four of 6 studies12,14,16,27 TRA, included adult patients and 2 enrolled only pediatric patients.13,26
Publications
Year
Study Design
Population
Karmiris et al
2009
Observational
168 CD
Hanauer et al (Classic 1 trial)
2006
299 CD
Chiu et al Classic 1/2
2013
Mazor et al
2013
Randomized, doubleblind, placebocontrolled Randomized, doubleblind, placebocontrolled Cross-sectional
Roblin et al
2013
Cross-sectional
Imaeda et al Bodini et al Velayos et al
2013 2012 2013
Dastrich et al
2011
Pradhan et al
2013
Sharma et al
2013
Sandborn et al
2012
Ferrante et al
2011
Open label
50 UC
West et al
2008
Retrospective study
30 CD
Regimen ADA induction: 160/80 (N ¼ 126), 80/40 (N ¼ 28); maintenance: 40 eow or ew ADA: 40/20 80/40, 160/80
276 CD
ADA: 40 eow; if flare: 40 ew
71 CD
Maintenance 40 mg eow
Follow-up Median, 20.4 (11.7–30) mo 4W
56W
NR duration of ADA: 9.8 mo NR duration of ADA 10.7 mo NR 2 yr NR
Impact of TRA in Clinical Response ADA cont. versus discont. TRA: W2, 10.4 versus 6.5 (P ¼ 0.02); W12, 9.3 versus 5.6 (P ¼ 0.02); W24, 9.9 versus 5.3 (P ¼ 0.04) TRA clinical remission versus no remission higher 8.1 versus 5.05 (P , 0.05) Clinical remission versus no TRA: W4, 5.42 versus 4.98 (P , 0.05); W24, 8.83 versus 5.82 (P , 0.05); W56, 9.37 versus 7.84 (P is nonsignificant) Correlation TRA/CR
40 IBD (22 CD) Cross-sectional 40 CD Prospective open label 22 CD Cross-sectional 54 IBD (52 CD) Open label 48 CD patients
Induction: 160/80; maintenance: 40 eow; if flare: 40 ew Maintenance Maintenance Maintenance
Open label (pediatric 181 CD patients) patients Double-blind 145 CD maintenance phase (pediatric patients) Randomized, double- 517 UC blind, control study
ADA induction phase 160/80 or 4W 80/40 based on body weight High dose based on body weight 40 52W mg or 20 mg eow low dose: 20 mg or 10 mg eow Adalimumab induction 160/80; W52 TRA at W8 and W52 higher in patients with clinical maintenance: 40 mg eow versus response under ADA placebo ADA induction 160/80; maintenance: TRA after escalation successful of escalation associated 40 mg eow with increase of TRA (P ¼ 0.023) ADA induction 160/80; maintenance: Median ¼ 318 d NR 40 eow or ew
ADA 40 eow on maintenance
NR
Inflamm Bowel Dis Volume 20, Number 7, July 2014
TABLE 1. Published Studies or Abstracts in Adults or Children on Adalimumab Trough Levels and Antibodies Against Adalimumab in CD and UC
TRA higher in clinical remission versus no 6.02 versus 3.2 (P ¼ 0.012) Correlation TRA/CRP TRA higher without flare 6.7 versus 2.1 (P , 0.01) TRA associated with no active disease versus active disease (P ¼ 0.01) Primary analysis deep remission TRA . 2 mg/mL; 84% with deep remission versus 70.8% without deep remission W4 response and remission rates were higher in patients with TRA above median Increase of rate of remission and response at W26 and 52 with TRA above the median concentration
1291
Pharmacokinetics of Adalimumab
www.ibdjournal.org |
| www.ibdjournal.org
Paul et al
1292
TABLE 1 (Continued) Publications
Impact of Current IS
Threshold Value of TRA
Positive AAA (%)
24% at baseline: no clinical impact TRA , 0.33: less clinical 9.2 or TRA impact time to dose benefit (Sen ¼ 95%, escalation longer 17 versus 12W PPV ¼ 81%) (P ¼ 0.008) Hanauer et al (Classic 31% during study No TRA threshold AUROC for 0.04 1 trial) clinical remission 0.54–0.62 No clinical or TRA impact Chiu et al Classic 1/2 N ¼ 93, TRA higher with current IS No TRA threshold 2.6 (P . 0.5) W24: 8.01 versus 5.96 W56: 8.62 versus 6.51 Mazor et al TRA cutoff for prediction of 30.5% 12.7% for clinical remission 5.85 AAA .3 mg/ (Sen ¼ 68%, Spe ¼ 70%, mL-eq LHR: 2.3) Roblin et al 12.5% no clinical or TRA cut-off for prediction of CR 22.5 pharmacological relevant 4.85 mg/mL (Sen ¼ 81%, Spe ¼ 67%, LHR ¼ 2.5) Imaeda et al TRA cutoff for prediction of 35 normal CRP, 5.9 Bodini et al NR Not done 9 22.3 Velayos et al NR TRA cut-off associated with CRP . 5 (P ¼ 0.001) Karmiris et al
NR NR NR NR NR NR
NR NR NR NR NR NR
NR NR NR NR NR 17
Impact of Current IS
Almost all patients (91.6%) displaying Do not increase AAA AAA had very low TRA and need development to stop adalimumab NR
NR
NR
NR
AAA inversely correlated with TRA (r ¼ 0.4111, P , 0.001 AAA positively related with disease activity (P , 0.001) AAA associated with clinical remission (LHR: 0.91; P ¼ 0.012)
No impact on AAA
No impact on AAA
Negative correlation with TRA and NR with CRP No impact on TRA and outcomes NR Positive AAA associated with NR elevated CRP (P ¼ 0.002); correlated with more active disease (P ¼ 0.01) NR NR NR NR NR NR NR NR NR NR AAA associated to nonresponse to AAA less frequent (7.7%) when ADA (OR ¼ 13.1; P ¼ 0.006) ADA is used in combotherapy, compared with monotherapy (20%), P . 0.5
ADA, adalimumab; CR, clinical remission; IS, immunosuppressant; LHR, lihehood ratio; PPV, positive predictive value; NR, not reported.
Inflamm Bowel Dis Volume 20, Number 7, July 2014
Dastrich et al Pradhan et al Sharma et al Sandborn et al Ferrante et al West et al
Impact of AAA
Inflamm Bowel Dis Volume 20, Number 7, July 2014
Pos/neg 1.12 3 NR 3 NR NR ELISA ELISA NA ELISA NA NR 5.9 5.85 Median 4.85 Median Clinician’s assessment CRP Clinician’s assessment PCDAI CDAI PCDAI Retrospective Retrospective Retrospective Prospetive Prospective Prospective Adult Adult Adult Pediatric Adult Pediatric 2008 2014 2013 2013 2014 2013 West et al Imaeda et al Mazor et al Pradhan et al Roblin et al Sharma et al
NR, not reported.
Remission Definition Methodology Population N Year
Four studies were included in the overall analysis,12,14,16,27 enrolling 163 CD patients. No studies evaluated AAA impact on clinical remission in pediatric patients. Three studies used the same ELISA method of measurement,12,14,16 the last one27 used an old radioimmunoassay. The cutoff of AAA ranged from 1.12 to 3 mg/mL in these 4 studies. The pooled OR for loss of clinical response to adalimumab in CD patients with AAAs was 10.15 (95% CI: 3.90–26.40, P , 0.0001) when compared with patients without AAAs. There was no heterogeneity among studies (Fig. 3; P ¼ 0.65). Visual inspection of funnel plot asymmetry showed no indication of publication bias. Both Begg’s and Egger’s tests were performed confirming the absence of publication bias (P ¼ 0.17 and 0.49, respectively). We next performed a sensitivity analysis excluding each study individually to determine whether there was any effect on heterogeneity or an impact on ORs. In
Study
Adalimumab Antibodies
TABLE 2. Characteristics of the Studies in the Meta-analysis
Five studies were included in the overall analysis12–14,16,26 enrolling 459 patients with CD. Patients with TRA over a predefined cutoff (4.85–5.9 mg/mL) were twice more likely to be in remission as compared with patients with TRA under this cutoff, with an OR of 2.6 (95% CI: 1.79–3.77, P , 0.0001). As heterogeneity existed between studies, 2 distinct meta-analyses were performed as the source of heterogeneity could be identified by separating studies evaluating adult patients12,14,16 from those evaluating pediatric population13,26 (N studies ¼ 3 and 2, respectively; Fig. 2). The association was stronger if the analysis was limited to the adult population (n ¼ 133) as compared with pediatric studies (n ¼ 326). Visual inspection of funnel plot asymmetry showed no indication of publication bias. Both Begg’s and Egger’s tests were performed, the latter one indicated a tendency toward publication bias (P ¼ 0.33 and 0.10, respectively). This could be explained by the absence of small trials with inconclusive results. To check if our results were robust, a sensitivity analysis was performed, excluding each study individually to determine whether there was any effect on heterogeneity or an impact on ORs: the exclusion of each study individually did not significantly modified the pooled ORs (pooled ORs ranging from 2.13 to 3.30).
Cut-off TRA (mg/mL)
Adalimumab Trough Levels
30 40 71 (118 samples) 181 22 145
TRA Assay Type
Cut-off AAA (mg/mL)
AAA Assay Type
Half of the included studies were prospective trials,12,13,26 and the other ones were of retrospective nature.14,16,27 The primary outcome measure was clinical remission,12,13,26 clinician’s global assessment,16,27 or biologic parameters such as CRP level.14 Characteristics of the studies are summarized in Table 2. All included studies except 2 controlled trials 13,26 had a high risk of bias. We first selected an article presenting post hoc analysis from the CLASSIC I and II trials.19 We contacted the authors to request further information, but could not gain access to the data, and this paper was thus excluded from our analysis. 19 As 5 of the 6 included studies did not involve patients with UC, only CD studies were included in the metaanalysis.
RIA ELISA ELISA NR ELISA NR
Pharmacokinetics of Adalimumab
www.ibdjournal.org |
1293
Inflamm Bowel Dis Volume 20, Number 7, July 2014
Paul et al
FIGURE 2. Forest plot of clinical response according to patients’ TRA over a predefined cutoff (4.85–5.9 mg/mL).
all cases, the exclusion of each study individually did not modify the pooled ORs (pooled ORs ranging from 8.87 to 13.23).
DISCUSSION Pharmacokinetics of anti-TNF therapy has important implications for patient management in IBD. Indeed, loss of response is frequent.29,30 Four anti-TNF agents are approved by FDA for the treatment of refractory IBD, namely infliximab, adalimumab, certolizumab pegol, and golimumab. Recently, some systematic reviews were published on the clinical utility of measuring infliximab trough levels and ATIs in IBD. We conducted the first systematic review and meta-analysis on the association of pharmacokinetics of adalimumab and clinical benefit in IBD. In our meta-analysis, the presence of AAA is associated with a significantly higher risk of loss of clinical response to adalimumab. High TRA are associated with clinical response in patients with CD. It is noteworthy that this association was stronger if the analysis was limited to the adult population with CD.
FIGURE 3. Forest plot of meta-analysis of “loss or response” according to patients’ AAA status.
1294
| www.ibdjournal.org
There are several strengths to this meta-analysis. First we conducted a systematic review to identify trials of interest. Furthermore, 489 patients were included. Finally, there is no publication bias, and source of heterogeneity in between-studies could be resolved by including only adult studies in a sensitivity analysis. This meta-analysis has some limitations. First, high risk of bias was present in the majority of included studies as 3 of the 6 were retrospective studies14,16,27 and only 2 controlled trials were conducted so far.13,26 Published studies on this topic lack uniform reporting of outcomes and cutoff. Finally, our analysis was restricted to patients with CD as only 1 study12 evaluated the impact of TRA and AAA on clinical response in UC. Overall, our results are comparable to data about the association of trough levels of infliximab and ATI and risk of loss of response to infliximab. Our meta-analysis could not analyze the impact of concomitant immunosuppressive therapy on disease outcome under adalimumab and their influence on TRA and AAA. We reported, however, that in 4 studies11,12,18,19 where the authors analyzed the impact of concomitant immunosuppressive therapy on disease outcome, it did not affect treatment efficacy and did not influence TRA.11,12,18,19 However, time to adalimumab dose escalation was longer under combination therapy (17 weeks versus 12 weeks; P ¼ 0.008).11 Chiu et al19 reported higher but not significantly different TRA at weeks 24 and 56 among patients on combination therapy. Randomized controlled trials addressing this issue are eagerly awaited. Interestingly, a cut-off value was identified in 4 studies, ranging from 4.8 to 5.9 mg/mL, whatever the method of measurement (3 ELISA different methods12,14,16 and 1 HMSA method).15 This cut-off value was finally comparable between these studies. In 1 study,11 patients with TRA ,0.33 mg/mL at least once during follow-up (sensitivity ¼ 95%, positive predictive value ¼ 81%) demonstrated significantly less sustained clinical benefit than those showing such low TRA (log-rank test; P ¼ 0.01). However, these findings should be interpreted cautiously because of small sample size (n ¼ 16). Importantly, the percentage of positive AAA varied strongly across studies between 0.04% and 35%.11,12,14–16,18,19,27,28 Differences between available assays, differing methodologies, and duration of follow-up may explain such discrepancy between studies. Almost all studies reported a negative correlation between AAA levels and TRA.11,12,14–16,27,28 Data are scarce on the association between AAA and intolerance to adalimumab. Karmiris et al11 concluded that side effect and injection site reaction were not explained by the presence of AAA. In a recent meta-analysis including 8 studies, the presence of ATIs was associated with a higher risk of acute infusion reactions, but not delayed hypersensitivity reactions, in patients with IBD.9 Concomitant immunomodulators reduced this risk. 9 More date are needed for adalimumab before drawing definite conclusions. In conclusion, similarly to infliximab, we could demonstrate that patients with TRA over a predefined cutoff were 2 times more likely to be in remission compared with patients with TRA under the cutoff, with an OR of 2.6. Moreover, the pooled
Inflamm Bowel Dis Volume 20, Number 7, July 2014
OR of loss of clinical response to adalimumab in patients with AAAs was 10.15 when compared with patients without AAAs in CD. Data are scarce in UC. Randomized controlled trials are awaited to further explore the clinical utility of measuring pharmacokinetics of adalimumab in IBD.
REFERENCES 1. Peyrin-Biroulet L, Deltenre P, de Suray N, et al. Efficacy and safety of tumor necrosis factor antagonists in Crohn’s disease: meta-analysis of placebo-controlled trials. Clin Gastroenterol Hepatol. 2008;6:644–653. 2. Rutgeerts P, Sandborn WJ, Feagan BG, et al. Infliximab for induction and maintenance therapy for ulcerative colitis. N Engl J Med. 2005;353: 2462–2476. 3. Sandborn WJ, van Assche G, Reinisch W, et al. Adalimumab induces and maintains clinical remission in patients with moderate-to-severe ulcerative colitis. Gastroenterology. 2012;142:257–265. 4. Sandborn WJ, Feagan BG, Marano C, et al. Subcutaneous golimumab induces clinical response and remission in patients with moderate-tosevere ulcerative colitis. Gastroenterology. 2014;146:85–95. 5. Sandborn WJ, Feagan BG, Marano C, et al. Subcutaneous golimumab maintains clinical response in patients with moderate-to-severe ulcerative colitis. Gastroenterology. 2014;146:96–109. 6. Colombel JF, Feagan BG, Sandborn WJ, et al. Therapeutic drug monitoring of biologics for inflammatory bowel disease. Inflamm Bowel Dis. 2012;18:349–358. 7. Lichtenstein GR. Letter from the editor. Gastroenterol Hepatol (N Y). 2013;9:67. 8. Nanda KS, Cheifetz AS, Moss AC. Impact of antibodies to infliximab on clinical outcomes and serum infliximab levels in patients with inflammatory bowel disease (IBD): a meta-analysis. Am J Gastroenterol. 2013;108: 40–47; quiz 48. 9. O’Meara S, Nanda KS, Moss AC. Antibodies to infliximab and risk of infusion reactions in patients with inflammatory bowel disease: a systematic review and meta-analysis. Inflamm Bowel Dis. 2014;20:1–6. 10. Ordas I, Feagan BG, Sandborn WJ. Therapeutic drug monitoring of tumor necrosis factor antagonists in inflammatory bowel disease. Clin Gastroenterol Hepatol. 2012;10:1079–1087; quiz e85–e86. 11. Karmiris K, Paintaud G, Noman M, et al. Influence of trough serum levels and immunogenicity on long-term outcome of adalimumab therapy in Crohn’s disease. Gastroenterology. 2009;137:1628–1640. 12. Roblin X, Marotte H, Rinaudo M, et al. Association between pharmacokinetics of adalimumab and mucosal healing in patients with inflammatory bowel diseases. Clin Gastroenterol Hepatol. 2014;12:80–84. 13. Pradhan RS, Sharma S, Thakkar R. Relationship between adalimumab concentrations and efficacy for the induction of clinical remission in pediatric patients with moderate to severe Crohn’s disease. Gastroenterology. 2013;144(Suppl 1):S230.
Pharmacokinetics of Adalimumab
14. Imaeda H, Takahashi K, Fujimoto T, et al. Clinical utility of newly developed immunoassays for serum concentrations of adalimumab and anti-adalimumab antibodies in patients with Crohn’s disease. J Gastroenterol. 2014;49:100–109. 15. Velayos FS, Sheibani S, Lockton S, et al. Prevalence of antibodies to adalimumab (ATA) and correlation between ATA and low serum drug concentration on CRP and clinical symptoms in a prospective sample of IBD patients. Gastroenterology. 2013;144(Suppl 1):S91. 16. Mazor Y, Koplov K, Ben Huret D. Evaluating Adalimumab Drug and Antibody Levels as Predictors of Clinical and Laboratory Response in Crohn’s Disease Patients. Vienna, Austria: ECCO; 2013:P517. 17. Dastrich P, Kozeluhova J, Bortlik M, et al. Infliximab but not adalimumab serum trough level is a useful predictor of deep remission in patients with Crohn’s disease. Gastroenterology. 2011;140(Suppl 1):S240. 18. Hanauer SB, Sandborn WJ, Rutgeerts P, et al. Human anti-tumor necrosis factor monoclonal antibody (adalimumab) in Crohn’s disease: the CLASSIC-I trial. Gastroenterology. 2006;130:323–333; quiz 591. 19. Chiu YL, Rubin DT, Vermeire S, et al. Serum adalimumab concentration and clinical remission in patients with Crohn’s disease. Inflamm Bowel Dis. 2013;19:1112–1122. 20. Guyatt G, Oxman AD, Akl EA, et al. GRADE guidelines: 1. IntroductionGRADE evidence profiles and summary of findings tables. J Clin Epidemiol. 2011;64:383–394. 21. Guyatt GH, Oxman AD, Vist G, et al. GRADE guidelines: 4. Rating the quality of evidence: study limitations (risk of bias). J Clin Epidemiol. 2011;64:407–415. 22. Higgins JP, Altman DG, Gotzsche PC, et al. The Cochrane Collaboration’s tool for assessing risk of bias in randomised trials. BMJ. 2011;343:d5928. 23. DerSimonian R, Laird N. Meta-analysis in clinical trials. Control Clin Trials. 1986;7:177–188. 24. Egger M, Davey Smith G, Schneider M, et al. Bias in meta-analysis detected by a simple, graphical test. BMJ. 1997;315:629–634. 25. Ferrante M, Karmiris K, Compernolle G, et al. Efficacy of ADA in Patients with UC: Restoration of Serum Levels After Dose Escalation Results in a Better Long Outcome. Stockolm, Sweden: UEGW; 2011:OP312. 26. Sharma S, Pradhan R, Thakkar R. Relationship between adalimumab concentration and efficacy for the maintenance of clinical remission in pediatric patients with moderate to severe Crohn’s disease. Gastroentrology. 2013; 144(Suppl 1):S231. 27. West RL, Zelinkova Z, Wolbink GJ, et al. Immunogenicity negatively influences the outcome of adalimumab treatment in Crohn’s disease. Aliment Pharmacol Ther. 2008;28:1122–1126. 28. Bodini G, Savarino V, Fazio V, et al. Relationship between drug serum concentration and clinical activity in patients with CD who achieved remission with adalimumab: a prospective study. Gastroenterology. 2012; 142(Suppl 1):S388. 29. Gisbert JP, Panes J. Loss of response and requirement of infliximab dose intensification in Crohn’s disease: a review. Am J Gastroenterol. 2009; 104:760–767. 30. Billioud V, Sandborn WJ, Peyrin-Biroulet L. Loss of response and need for adalimumab dose intensification in Crohn’s disease: a systematic review. Am J Gastroenterol. 2011;106:674–684.
www.ibdjournal.org |
1295
Pharmacokinetics of Adalimumab in Inflammatory Bowel Diseases: A Systematic Review and Meta-analysis Stephane Paul, PhD,* Amelie Carla Moreau, PhD,* Emilie Del Tedesco, MD,† Melanie Rinaudo, MD, PhD,* Jean-Marc Phelip, MD, PhD,† Christian Genin, MD, PhD,* Laurent Peyrin-Biroulet, MD, PhD,‡ and Xavier Roblin, MD, PhD†
Background: The aim of this meta-analysis was to explore the magnitude of the association between pharmacokinetics of adalimumab and clinical response in patients with inflammatory bowel disease. Methods: A literature search was performed up to December 2013. MEDLINE, EMBASE, Cochrane, and meeting abstracts were reviewed. Studies were included if they analyzed the association of trough levels of adalimumab (TRA) or antibodies against adalimumab (AAA) with clinical response in adult or pediatric inflammatory bowel disease. A Mantel–Haenszel pooled risk estimate provided a measure of the association.
Results: Fourteen studies enrolling 1941 patients with inflammatory bowel disease were included in the systematic review. Thirteen studies analyzed clinical outcomes according to TRA. In only 1 study, there was no correlation between high TRA and clinical response. Six of the 7 studies reported a negative correlation between AAA and clinical outcomes. Six studies enrolling 536 patients (Crohn’s disease [CD] only) met the meta-analysis inclusion criteria. The pooled odds ratio (OR) for loss of clinical response to adalimumab in patients with CD (N ¼ 4) with positive AAAs was 10.15 (95% confidence interval [CI]: 3.90–26.40, P , 0.0001). Patients with CD with TRA over a predefined cutoff were more likely to be in clinical remission with an OR of 2.6 (95% CI: 1.79–3.77, P , 0.0001). The association was stronger if the analysis was limited to the adult population (N ¼ 3, OR: 7.05, 95% CI: 3.58–13.9, P , 0.0001). Conclusions: The presence of AAA is associated with a higher risk of loss of clinical response to adalimumab, whereas high TRA is associated with greater clinical response rates in CD. More data are needed in ulcerative colitis. (Inflamm Bowel Dis 2014;20:1288–1295) Key Words: inflammatory bowel diseases, adalimumab trough levels, antibodies to adalimumab, therapeutic drug monitoring, clinical response
T
hree anti-tumor necrosis factor (TNF) agents, namely, infliximab, adalimumab, and certolizumab pegol, are approved by the Food and Drug Administration (FDA) for the treatment of refractory luminal Crohn’s disease (CD).1 Infliximab, adalimumab, and golimumab are approved by FDA for the treatment of refractory ulcerative colitis (UC).2–5 Measuring antibodies against anti-TNF therapy and drug trough levels are increasingly used to improve disease outcomes in inflammatory bowel diseases (IBD).6,7 In a meta-analysis including 13 studies and 1378 patients with IBD,8 the presence of antibodies against infliximab (ATIs) was associated with a significantly higher risk of loss of clinical response to infliximab {relative risk of loss of Received for publication February 1, 2014; Accepted February 27, 2014. From the *Laboratoire d’Immunologie et d’Immunomonitoring, CHU de SaintEtienne, Saint-Etienne, France; †Service de Gastrologie-Entérologie-Hépatologie, CHU de Saint-Etienne, Saint-Etienne, France; and ‡Department of Gastroenterology, Inserm U954, University Hospital of Nancy, Université de Lorraine, Nancy, France. L. Peyrin-Biroulet received Lecture and/or consulting fees from MSD, Abbvie, UCB-pharma, and X. Roblin received Lecture and consulting fees from Merck, Abbvie, and Theradiag. The remaining authors have no conflicts of interest to disclose. Reprints: Xavier Roblin, MD, PhD, Department of Gastroenterology, CHU de Saint-Etienne, Avenue Albert Raymond, Saint-Etienne 42023, France (e-mail: xavier. [email protected]). Copyright © 2014 Crohn’s & Colitis Foundation of America, Inc. DOI 10.1097/MIB.0000000000000037 Published online 14 May 2014.
1288
| www.ibdjournal.org
response of 2.2 (95% confidence interval [CI]: 0.5–9.0, P ¼ 0.3)} and lower serum infliximab levels (trough levels of infliximab). In a second meta-analysis including 8 studies, the presence of ATIs was associated with a higher risk of acute infusion reactions, but not delayed hypersensitivity reactions, in patients with IBD.9 Concomitant immunomodulators use reduced this risk.9 Based on these findings, Ordas et al10 concluded that incorporation of therapeutic drug monitoring into clinical practice may allow clinicians to optimize treatment by maintaining effective drug concentrations over time. Adalimumab is a fully human anti-TNF. Recently, several studies investigated the association of pharmacokinetics of adalimumab with clinical benefits in IBD.11–18 However, some studies were of small sample size, and some of them provided conflicting results.11,12,19 Our aim was first to review available evidence regarding the association between pharmacokinetics of adalimumab and clinical response in IBD before conducting a meta-analysis of studies fulfilling inclusion criteria.
METHODS Study Selection We conducted a comprehensive computerized search for English and non-English language publications listed in the electronic databases of MEDLINE (source PUBMED, 1966 to Inflamm Bowel Dis Volume 20, Number 7, July 2014
Inflamm Bowel Dis Volume 20, Number 7, July 2014
December 2013), the Cochrane Library (Issue 12, 2013), and EMBASE (1980 to December 2013). We searched and combined the following terms: “drug monitoring, inflammatory bowel disease, Crohn’s disease, ulcerative colitis, adalimumab, drug level, antibodies, outcomes, treatment response, impact of drug level, impact of antibody formation.” We also hand-searched abstracts from the annual meetings of Digestive Disease Week and the United European Gastroenterology Week between 2005 and 2013, as well as references from review articles, meta-analysis studies, and published randomized controlled trials so as to identify additional articles. Authors of identified studies and pharmaceutical companies were also contacted to retrieve missing or unpublished data. Studies were selected based on the title and/or the abstract by 2 researchers (A.C.M. and X.R.). The reviewers checked whether inclusion and exclusion criteria were met and, in the event of uncertainty, the full text of articles was retrieved and reviewed. To be included in the systematic review, the focus of the study had to be the reporting of clinical response under adalimumab. We included all studies (controlled trials, observational studies, and cohort studies) that reported antibodies against adalimumab (AAA) and/or trough levels of adalimumab (TRA) and clinical outcomes in patients with IBD who were treated with adalimumab. Where AAA was studied, the primary outcome measure was the loss of response, defined as relapse of clinical symptoms in patients who were in clinical remission from, or had responded to, adalimumab. No prespecified score was used to determine this outcome, as we expected there would be no universal definition of the outcome across the studies. Where TRA were studied, the primary outcome measure was clinical response. Studies were excluded if they (1) were review articles, (2) the population presented was non-IBD patients, (3) did not measure either TRA or AAA, (4) did not report clinical outcomes, and (5) were studies using other anti-TNFs (infliximab or certolizumab). If there was any suspicion of a duplicate study, the most recent study with the largest cohort was considered for inclusion. There were additional inclusion criteria for the metaanalysis. Studies were included if they reported clinical outcomes according to AAA status or TRA with a predefined cutoff, thus allowing the construction of 2 · 2 tables.
Outcome Assessment Outcome measures were defined a priori. The efficacy end points were clinical remission, absence of flare-up, deep remission, or normalization of C-reactive protein (CRP). The definition of clinical activity varied across studies (Table 2).
Data Extraction Data abstraction was carried out independently by 2 investigators (A.C.M. and X.R.) using standardized data collection forms. If needed, authors were contacted regarding specific questions relating to their study. The independent reviewers conferred after data extraction was complete, discrepancies were identified, and review of the relevant article with a third author (S.P.) led to consensus.
Pharmacokinetics of Adalimumab
The following data were retrieved where possible: journal, year, author, number of patients in the study, method of selection of cohorts, drug regimen, assay type used to measure TRA and/or AAA, numbers of AAA-positive and AAA-negative patients, numbers of TRA over the cutoff and under the cutoff, methods of measurement of clinical outcomes, and reporting of outcomes used. Assessment of quality of randomized controlled trials and observational studies was performed using The GRADE (Grading of Recommendations Assessment, Development and Evaluation) criteria.20 For observational studies, methodological quality was assessed by determining the eligibility criteria, degree of measurement of both exposure and outcome, control for confounders, and completeness of follow-up.21
Data Synthesis and Statistical Analysis The primary goal of the meta-analysis was to assess the association between AAA or adalimumab trough levels and clinical remission in patients with IBD. Secondary objectives were to evaluate whether or not the association was influenced by differences in study design such as the presence of a selection and/ or information bias. A Mantel–Haenszel pooled risk estimate provided a measure of that association. Raw data from included studies (absolute numbers) were used to construct 2 · 2 contingency tables, and unadjusted pooled odd ratios (ORs) with 95% CIs were calculated using Review Manager (RevMan 5.1; Nordic Cochrane Centre, Copenhagen, Denmark) for dichotomous outcomes. Forest plots and funnels plots were generated using R version 3.0.2 (R, University of Auckland, Auckland, New Zealand). The Q test was used to assess for heterogeneity and I2 statistic to quantify the percentage of heterogeneity because of between-study variations; a value of P , 0.10 was considered statistically significant.22 Any heterogeneity identified prompted subgroup analysis in an attempt to explain these findings. If needed, a random effects model by DerSimonian and Laird23 was used to account for variations between studies and gives a more conservative pooled estimate in case where heterogeneity between studies could not be explained. A funnel plot was generated and checked for evidence of asymmetry, and therefore possible publication bias or other small study effects, using the Egger’s test, a P value ,0.10 was regarded as significant.24 Sensitivity analyses were performed for all outcomes where 3 or more studies were included. Sensitivity analysis included sensitivity based on individual studies, sample size, drug regimen, length of the study, outcome measure, and drugs used in the control arm. Significance levels were set at a P value of ,0.05.
RESULTS Systematic Review of the Association Between Pharmacokinetics of Adalimumab and Clinical Benefit Database searches identified 499 references, and 14 articles were included in the systematic review (Fig. 1): 11 studies in CD, www.ibdjournal.org |
1289
Inflamm Bowel Dis Volume 20, Number 7, July 2014
Paul et al
2 in UC,3,25 and 1 included both CD and UC patients12 (Table 1). Two studies included patients with pediatric CD.13,26 Twelve studies included adult patients with IBD.3,11,12,14–19,25,27,28 Six studies evaluated disease activity in patients on adalimumab treatment according to TRA and AAA,11,12,14–16,28 7 studies analyzed disease activity according to TRA with no reported data on AAA,3,13,17–19,25,26 and 1 study analyzed disease activity according to AAA without data on TRA.27 Four randomized controlled trials,3,18,19,26 4 prospective crosssectional studies,12,14–16 5 observational studies,11,13,17,25,28 and 1 retrospective study17 were included in the systematic review. A quiescent disease was defined by clinical remission in 10 studies11–16,18,19,25,28 by CRP normalization in 1 study,14 by deep remission defined by a clinical remission and CRP normalization17 in 1 study, and 1 study evaluated clinical response after adalimumab dose escalation.25 One study analyzed the correlation between clinical response and AAA without data on TRA in CD.27
was performed in the one by Sandborn et al3 Ferrante et al25 analyzed TRA after adalimumab dose escalation, and Roblin et al12 analyzed the association between TRA and clinical remission by pooling UC and CD. Ten CD11,13–19,26,28 studies looked at the association between TRA and clinical response. Only 1 study reported no increased rate of clinical response among CD patients with higher TRA.19
Identification of an Optimal Cutoff for TRA for Predicting Clinical Remission Area under curve was used to identify the optimal cutoff for TRA for predicting clinical remission in 7 studies.11,12,14–16,18,19 A cutoff value was identified in 4 studies, ranging from 4.8 to 5.9 mg/ mL.12,14–16 For Karmiris et al,11 patients who had TRA ,0.33 mg/mL at least once (sensitivity, 95%; positive predictive value, 81%) demonstrated significantly less sustained clinical benefit than patients never showing such low trough serum concentration (log-rank test; P , 0.01).
Pediatric Versus Adult Studies Two pediatric studies (N ¼ 326) included only CD patients.13,26 In both studies, there was an increased rate of remission and response among patients with TRA above the median concentration.13,26 No data about AAA were reported in these 2 studies.13,26 In the adult population (N ¼ 1615 IBD patients), there was a correlation between high TRA and clinical response in all 11 studies3,11,12,14–18,25,28 except the study by Chiu et al19 (Table 1).
Impact of Concomitant Immunosuppressive Treatment
Crohn’s Disease Versus Ulcerative Colitis
In 4 studies11,12,18,19 where the authors analyzed the impact of concomitant immunosuppressive therapy on disease outcome, it did not affect treatment efficacy and did not influence TRA.11,12,18,19 However, time to adalimumab dose escalation was longer under combination therapy (17 weeks versus 12 weeks; P ¼ 0.008).8 Chiu et al19 reported higher but not significantly different TRA at weeks 24 and 56 among patients on combination therapy.
Two studies included only UC patients3,25 and 1 combined UC and CD patients.12 All 3 studies found an association between TRA and clinical response, even though no statistical comparison
Impact of Antibodies Against Adalimumab on Disease Outcomes Nine studies investigated so far the percentage of positive AAA among adalimumab-treated IBD patients,11,12,14–16,18,19,27,28 7 studies11,12,14–16,27,28 have been published as full-length papers and all of them included only adult patients. The percentage of positive AAA varied between 0.04% and 35%.11,12,14–16,18,19,27,28 Differences between available assays and differing methodologies may explain these variations. Almost all studies reported a negative correlation between AAA levels and TRA.11,12,14–16,27,28 Six of the 7 studies found a negative correlation between AAA and clinical outcomes (5 with clinical disease activity11,12,15,16,27 and 2 with CRP levels).14,15 These results were not confirmed in 1 study.28 Concomitant immunosuppressive therapy did not affect treatment outcome and did not influence AAA formation.11,12,16,27
Meta-analysis of the Association Between Pharmacokinetics of Adalimumab and Clinical Benefit
FIGURE 1. Flow chart of study selection.
1290
| www.ibdjournal.org
Of the 14 studies selected and included in the systematic review, only 6 studies met the inclusion criteria and were included in the meta-analysis.12–14,16,26,27 Three studies evaluated both TRA and AAA,12,14,16 2 only 13,26 and 1 study only AAA.27 Four of 6 studies12,14,16,27 TRA, included adult patients and 2 enrolled only pediatric patients.13,26
Publications
Year
Study Design
Population
Karmiris et al
2009
Observational
168 CD
Hanauer et al (Classic 1 trial)
2006
299 CD
Chiu et al Classic 1/2
2013
Mazor et al
2013
Randomized, doubleblind, placebocontrolled Randomized, doubleblind, placebocontrolled Cross-sectional
Roblin et al
2013
Cross-sectional
Imaeda et al Bodini et al Velayos et al
2013 2012 2013
Dastrich et al
2011
Pradhan et al
2013
Sharma et al
2013
Sandborn et al
2012
Ferrante et al
2011
Open label
50 UC
West et al
2008
Retrospective study
30 CD
Regimen ADA induction: 160/80 (N ¼ 126), 80/40 (N ¼ 28); maintenance: 40 eow or ew ADA: 40/20 80/40, 160/80
276 CD
ADA: 40 eow; if flare: 40 ew
71 CD
Maintenance 40 mg eow
Follow-up Median, 20.4 (11.7–30) mo 4W
56W
NR duration of ADA: 9.8 mo NR duration of ADA 10.7 mo NR 2 yr NR
Impact of TRA in Clinical Response ADA cont. versus discont. TRA: W2, 10.4 versus 6.5 (P ¼ 0.02); W12, 9.3 versus 5.6 (P ¼ 0.02); W24, 9.9 versus 5.3 (P ¼ 0.04) TRA clinical remission versus no remission higher 8.1 versus 5.05 (P , 0.05) Clinical remission versus no TRA: W4, 5.42 versus 4.98 (P , 0.05); W24, 8.83 versus 5.82 (P , 0.05); W56, 9.37 versus 7.84 (P is nonsignificant) Correlation TRA/CR
40 IBD (22 CD) Cross-sectional 40 CD Prospective open label 22 CD Cross-sectional 54 IBD (52 CD) Open label 48 CD patients
Induction: 160/80; maintenance: 40 eow; if flare: 40 ew Maintenance Maintenance Maintenance
Open label (pediatric 181 CD patients) patients Double-blind 145 CD maintenance phase (pediatric patients) Randomized, double- 517 UC blind, control study
ADA induction phase 160/80 or 4W 80/40 based on body weight High dose based on body weight 40 52W mg or 20 mg eow low dose: 20 mg or 10 mg eow Adalimumab induction 160/80; W52 TRA at W8 and W52 higher in patients with clinical maintenance: 40 mg eow versus response under ADA placebo ADA induction 160/80; maintenance: TRA after escalation successful of escalation associated 40 mg eow with increase of TRA (P ¼ 0.023) ADA induction 160/80; maintenance: Median ¼ 318 d NR 40 eow or ew
ADA 40 eow on maintenance
NR
Inflamm Bowel Dis Volume 20, Number 7, July 2014
TABLE 1. Published Studies or Abstracts in Adults or Children on Adalimumab Trough Levels and Antibodies Against Adalimumab in CD and UC
TRA higher in clinical remission versus no 6.02 versus 3.2 (P ¼ 0.012) Correlation TRA/CRP TRA higher without flare 6.7 versus 2.1 (P , 0.01) TRA associated with no active disease versus active disease (P ¼ 0.01) Primary analysis deep remission TRA . 2 mg/mL; 84% with deep remission versus 70.8% without deep remission W4 response and remission rates were higher in patients with TRA above median Increase of rate of remission and response at W26 and 52 with TRA above the median concentration
1291
Pharmacokinetics of Adalimumab
www.ibdjournal.org |
| www.ibdjournal.org
Paul et al
1292
TABLE 1 (Continued) Publications
Impact of Current IS
Threshold Value of TRA
Positive AAA (%)
24% at baseline: no clinical impact TRA , 0.33: less clinical 9.2 or TRA impact time to dose benefit (Sen ¼ 95%, escalation longer 17 versus 12W PPV ¼ 81%) (P ¼ 0.008) Hanauer et al (Classic 31% during study No TRA threshold AUROC for 0.04 1 trial) clinical remission 0.54–0.62 No clinical or TRA impact Chiu et al Classic 1/2 N ¼ 93, TRA higher with current IS No TRA threshold 2.6 (P . 0.5) W24: 8.01 versus 5.96 W56: 8.62 versus 6.51 Mazor et al TRA cutoff for prediction of 30.5% 12.7% for clinical remission 5.85 AAA .3 mg/ (Sen ¼ 68%, Spe ¼ 70%, mL-eq LHR: 2.3) Roblin et al 12.5% no clinical or TRA cut-off for prediction of CR 22.5 pharmacological relevant 4.85 mg/mL (Sen ¼ 81%, Spe ¼ 67%, LHR ¼ 2.5) Imaeda et al TRA cutoff for prediction of 35 normal CRP, 5.9 Bodini et al NR Not done 9 22.3 Velayos et al NR TRA cut-off associated with CRP . 5 (P ¼ 0.001) Karmiris et al
NR NR NR NR NR NR
NR NR NR NR NR NR
NR NR NR NR NR 17
Impact of Current IS
Almost all patients (91.6%) displaying Do not increase AAA AAA had very low TRA and need development to stop adalimumab NR
NR
NR
NR
AAA inversely correlated with TRA (r ¼ 0.4111, P , 0.001 AAA positively related with disease activity (P , 0.001) AAA associated with clinical remission (LHR: 0.91; P ¼ 0.012)
No impact on AAA
No impact on AAA
Negative correlation with TRA and NR with CRP No impact on TRA and outcomes NR Positive AAA associated with NR elevated CRP (P ¼ 0.002); correlated with more active disease (P ¼ 0.01) NR NR NR NR NR NR NR NR NR NR AAA associated to nonresponse to AAA less frequent (7.7%) when ADA (OR ¼ 13.1; P ¼ 0.006) ADA is used in combotherapy, compared with monotherapy (20%), P . 0.5
ADA, adalimumab; CR, clinical remission; IS, immunosuppressant; LHR, lihehood ratio; PPV, positive predictive value; NR, not reported.
Inflamm Bowel Dis Volume 20, Number 7, July 2014
Dastrich et al Pradhan et al Sharma et al Sandborn et al Ferrante et al West et al
Impact of AAA
Inflamm Bowel Dis Volume 20, Number 7, July 2014
Pos/neg 1.12 3 NR 3 NR NR ELISA ELISA NA ELISA NA NR 5.9 5.85 Median 4.85 Median Clinician’s assessment CRP Clinician’s assessment PCDAI CDAI PCDAI Retrospective Retrospective Retrospective Prospetive Prospective Prospective Adult Adult Adult Pediatric Adult Pediatric 2008 2014 2013 2013 2014 2013 West et al Imaeda et al Mazor et al Pradhan et al Roblin et al Sharma et al
NR, not reported.
Remission Definition Methodology Population N Year
Four studies were included in the overall analysis,12,14,16,27 enrolling 163 CD patients. No studies evaluated AAA impact on clinical remission in pediatric patients. Three studies used the same ELISA method of measurement,12,14,16 the last one27 used an old radioimmunoassay. The cutoff of AAA ranged from 1.12 to 3 mg/mL in these 4 studies. The pooled OR for loss of clinical response to adalimumab in CD patients with AAAs was 10.15 (95% CI: 3.90–26.40, P , 0.0001) when compared with patients without AAAs. There was no heterogeneity among studies (Fig. 3; P ¼ 0.65). Visual inspection of funnel plot asymmetry showed no indication of publication bias. Both Begg’s and Egger’s tests were performed confirming the absence of publication bias (P ¼ 0.17 and 0.49, respectively). We next performed a sensitivity analysis excluding each study individually to determine whether there was any effect on heterogeneity or an impact on ORs. In
Study
Adalimumab Antibodies
TABLE 2. Characteristics of the Studies in the Meta-analysis
Five studies were included in the overall analysis12–14,16,26 enrolling 459 patients with CD. Patients with TRA over a predefined cutoff (4.85–5.9 mg/mL) were twice more likely to be in remission as compared with patients with TRA under this cutoff, with an OR of 2.6 (95% CI: 1.79–3.77, P , 0.0001). As heterogeneity existed between studies, 2 distinct meta-analyses were performed as the source of heterogeneity could be identified by separating studies evaluating adult patients12,14,16 from those evaluating pediatric population13,26 (N studies ¼ 3 and 2, respectively; Fig. 2). The association was stronger if the analysis was limited to the adult population (n ¼ 133) as compared with pediatric studies (n ¼ 326). Visual inspection of funnel plot asymmetry showed no indication of publication bias. Both Begg’s and Egger’s tests were performed, the latter one indicated a tendency toward publication bias (P ¼ 0.33 and 0.10, respectively). This could be explained by the absence of small trials with inconclusive results. To check if our results were robust, a sensitivity analysis was performed, excluding each study individually to determine whether there was any effect on heterogeneity or an impact on ORs: the exclusion of each study individually did not significantly modified the pooled ORs (pooled ORs ranging from 2.13 to 3.30).
Cut-off TRA (mg/mL)
Adalimumab Trough Levels
30 40 71 (118 samples) 181 22 145
TRA Assay Type
Cut-off AAA (mg/mL)
AAA Assay Type
Half of the included studies were prospective trials,12,13,26 and the other ones were of retrospective nature.14,16,27 The primary outcome measure was clinical remission,12,13,26 clinician’s global assessment,16,27 or biologic parameters such as CRP level.14 Characteristics of the studies are summarized in Table 2. All included studies except 2 controlled trials 13,26 had a high risk of bias. We first selected an article presenting post hoc analysis from the CLASSIC I and II trials.19 We contacted the authors to request further information, but could not gain access to the data, and this paper was thus excluded from our analysis. 19 As 5 of the 6 included studies did not involve patients with UC, only CD studies were included in the metaanalysis.
RIA ELISA ELISA NR ELISA NR
Pharmacokinetics of Adalimumab
www.ibdjournal.org |
1293
Inflamm Bowel Dis Volume 20, Number 7, July 2014
Paul et al
FIGURE 2. Forest plot of clinical response according to patients’ TRA over a predefined cutoff (4.85–5.9 mg/mL).
all cases, the exclusion of each study individually did not modify the pooled ORs (pooled ORs ranging from 8.87 to 13.23).
DISCUSSION Pharmacokinetics of anti-TNF therapy has important implications for patient management in IBD. Indeed, loss of response is frequent.29,30 Four anti-TNF agents are approved by FDA for the treatment of refractory IBD, namely infliximab, adalimumab, certolizumab pegol, and golimumab. Recently, some systematic reviews were published on the clinical utility of measuring infliximab trough levels and ATIs in IBD. We conducted the first systematic review and meta-analysis on the association of pharmacokinetics of adalimumab and clinical benefit in IBD. In our meta-analysis, the presence of AAA is associated with a significantly higher risk of loss of clinical response to adalimumab. High TRA are associated with clinical response in patients with CD. It is noteworthy that this association was stronger if the analysis was limited to the adult population with CD.
FIGURE 3. Forest plot of meta-analysis of “loss or response” according to patients’ AAA status.
1294
| www.ibdjournal.org
There are several strengths to this meta-analysis. First we conducted a systematic review to identify trials of interest. Furthermore, 489 patients were included. Finally, there is no publication bias, and source of heterogeneity in between-studies could be resolved by including only adult studies in a sensitivity analysis. This meta-analysis has some limitations. First, high risk of bias was present in the majority of included studies as 3 of the 6 were retrospective studies14,16,27 and only 2 controlled trials were conducted so far.13,26 Published studies on this topic lack uniform reporting of outcomes and cutoff. Finally, our analysis was restricted to patients with CD as only 1 study12 evaluated the impact of TRA and AAA on clinical response in UC. Overall, our results are comparable to data about the association of trough levels of infliximab and ATI and risk of loss of response to infliximab. Our meta-analysis could not analyze the impact of concomitant immunosuppressive therapy on disease outcome under adalimumab and their influence on TRA and AAA. We reported, however, that in 4 studies11,12,18,19 where the authors analyzed the impact of concomitant immunosuppressive therapy on disease outcome, it did not affect treatment efficacy and did not influence TRA.11,12,18,19 However, time to adalimumab dose escalation was longer under combination therapy (17 weeks versus 12 weeks; P ¼ 0.008).11 Chiu et al19 reported higher but not significantly different TRA at weeks 24 and 56 among patients on combination therapy. Randomized controlled trials addressing this issue are eagerly awaited. Interestingly, a cut-off value was identified in 4 studies, ranging from 4.8 to 5.9 mg/mL, whatever the method of measurement (3 ELISA different methods12,14,16 and 1 HMSA method).15 This cut-off value was finally comparable between these studies. In 1 study,11 patients with TRA ,0.33 mg/mL at least once during follow-up (sensitivity ¼ 95%, positive predictive value ¼ 81%) demonstrated significantly less sustained clinical benefit than those showing such low TRA (log-rank test; P ¼ 0.01). However, these findings should be interpreted cautiously because of small sample size (n ¼ 16). Importantly, the percentage of positive AAA varied strongly across studies between 0.04% and 35%.11,12,14–16,18,19,27,28 Differences between available assays, differing methodologies, and duration of follow-up may explain such discrepancy between studies. Almost all studies reported a negative correlation between AAA levels and TRA.11,12,14–16,27,28 Data are scarce on the association between AAA and intolerance to adalimumab. Karmiris et al11 concluded that side effect and injection site reaction were not explained by the presence of AAA. In a recent meta-analysis including 8 studies, the presence of ATIs was associated with a higher risk of acute infusion reactions, but not delayed hypersensitivity reactions, in patients with IBD.9 Concomitant immunomodulators reduced this risk. 9 More date are needed for adalimumab before drawing definite conclusions. In conclusion, similarly to infliximab, we could demonstrate that patients with TRA over a predefined cutoff were 2 times more likely to be in remission compared with patients with TRA under the cutoff, with an OR of 2.6. Moreover, the pooled
Inflamm Bowel Dis Volume 20, Number 7, July 2014
OR of loss of clinical response to adalimumab in patients with AAAs was 10.15 when compared with patients without AAAs in CD. Data are scarce in UC. Randomized controlled trials are awaited to further explore the clinical utility of measuring pharmacokinetics of adalimumab in IBD.
REFERENCES 1. Peyrin-Biroulet L, Deltenre P, de Suray N, et al. Efficacy and safety of tumor necrosis factor antagonists in Crohn’s disease: meta-analysis of placebo-controlled trials. Clin Gastroenterol Hepatol. 2008;6:644–653. 2. Rutgeerts P, Sandborn WJ, Feagan BG, et al. Infliximab for induction and maintenance therapy for ulcerative colitis. N Engl J Med. 2005;353: 2462–2476. 3. Sandborn WJ, van Assche G, Reinisch W, et al. Adalimumab induces and maintains clinical remission in patients with moderate-to-severe ulcerative colitis. Gastroenterology. 2012;142:257–265. 4. Sandborn WJ, Feagan BG, Marano C, et al. Subcutaneous golimumab induces clinical response and remission in patients with moderate-tosevere ulcerative colitis. Gastroenterology. 2014;146:85–95. 5. Sandborn WJ, Feagan BG, Marano C, et al. Subcutaneous golimumab maintains clinical response in patients with moderate-to-severe ulcerative colitis. Gastroenterology. 2014;146:96–109. 6. Colombel JF, Feagan BG, Sandborn WJ, et al. Therapeutic drug monitoring of biologics for inflammatory bowel disease. Inflamm Bowel Dis. 2012;18:349–358. 7. Lichtenstein GR. Letter from the editor. Gastroenterol Hepatol (N Y). 2013;9:67. 8. Nanda KS, Cheifetz AS, Moss AC. Impact of antibodies to infliximab on clinical outcomes and serum infliximab levels in patients with inflammatory bowel disease (IBD): a meta-analysis. Am J Gastroenterol. 2013;108: 40–47; quiz 48. 9. O’Meara S, Nanda KS, Moss AC. Antibodies to infliximab and risk of infusion reactions in patients with inflammatory bowel disease: a systematic review and meta-analysis. Inflamm Bowel Dis. 2014;20:1–6. 10. Ordas I, Feagan BG, Sandborn WJ. Therapeutic drug monitoring of tumor necrosis factor antagonists in inflammatory bowel disease. Clin Gastroenterol Hepatol. 2012;10:1079–1087; quiz e85–e86. 11. Karmiris K, Paintaud G, Noman M, et al. Influence of trough serum levels and immunogenicity on long-term outcome of adalimumab therapy in Crohn’s disease. Gastroenterology. 2009;137:1628–1640. 12. Roblin X, Marotte H, Rinaudo M, et al. Association between pharmacokinetics of adalimumab and mucosal healing in patients with inflammatory bowel diseases. Clin Gastroenterol Hepatol. 2014;12:80–84. 13. Pradhan RS, Sharma S, Thakkar R. Relationship between adalimumab concentrations and efficacy for the induction of clinical remission in pediatric patients with moderate to severe Crohn’s disease. Gastroenterology. 2013;144(Suppl 1):S230.
Pharmacokinetics of Adalimumab
14. Imaeda H, Takahashi K, Fujimoto T, et al. Clinical utility of newly developed immunoassays for serum concentrations of adalimumab and anti-adalimumab antibodies in patients with Crohn’s disease. J Gastroenterol. 2014;49:100–109. 15. Velayos FS, Sheibani S, Lockton S, et al. Prevalence of antibodies to adalimumab (ATA) and correlation between ATA and low serum drug concentration on CRP and clinical symptoms in a prospective sample of IBD patients. Gastroenterology. 2013;144(Suppl 1):S91. 16. Mazor Y, Koplov K, Ben Huret D. Evaluating Adalimumab Drug and Antibody Levels as Predictors of Clinical and Laboratory Response in Crohn’s Disease Patients. Vienna, Austria: ECCO; 2013:P517. 17. Dastrich P, Kozeluhova J, Bortlik M, et al. Infliximab but not adalimumab serum trough level is a useful predictor of deep remission in patients with Crohn’s disease. Gastroenterology. 2011;140(Suppl 1):S240. 18. Hanauer SB, Sandborn WJ, Rutgeerts P, et al. Human anti-tumor necrosis factor monoclonal antibody (adalimumab) in Crohn’s disease: the CLASSIC-I trial. Gastroenterology. 2006;130:323–333; quiz 591. 19. Chiu YL, Rubin DT, Vermeire S, et al. Serum adalimumab concentration and clinical remission in patients with Crohn’s disease. Inflamm Bowel Dis. 2013;19:1112–1122. 20. Guyatt G, Oxman AD, Akl EA, et al. GRADE guidelines: 1. IntroductionGRADE evidence profiles and summary of findings tables. J Clin Epidemiol. 2011;64:383–394. 21. Guyatt GH, Oxman AD, Vist G, et al. GRADE guidelines: 4. Rating the quality of evidence: study limitations (risk of bias). J Clin Epidemiol. 2011;64:407–415. 22. Higgins JP, Altman DG, Gotzsche PC, et al. The Cochrane Collaboration’s tool for assessing risk of bias in randomised trials. BMJ. 2011;343:d5928. 23. DerSimonian R, Laird N. Meta-analysis in clinical trials. Control Clin Trials. 1986;7:177–188. 24. Egger M, Davey Smith G, Schneider M, et al. Bias in meta-analysis detected by a simple, graphical test. BMJ. 1997;315:629–634. 25. Ferrante M, Karmiris K, Compernolle G, et al. Efficacy of ADA in Patients with UC: Restoration of Serum Levels After Dose Escalation Results in a Better Long Outcome. Stockolm, Sweden: UEGW; 2011:OP312. 26. Sharma S, Pradhan R, Thakkar R. Relationship between adalimumab concentration and efficacy for the maintenance of clinical remission in pediatric patients with moderate to severe Crohn’s disease. Gastroentrology. 2013; 144(Suppl 1):S231. 27. West RL, Zelinkova Z, Wolbink GJ, et al. Immunogenicity negatively influences the outcome of adalimumab treatment in Crohn’s disease. Aliment Pharmacol Ther. 2008;28:1122–1126. 28. Bodini G, Savarino V, Fazio V, et al. Relationship between drug serum concentration and clinical activity in patients with CD who achieved remission with adalimumab: a prospective study. Gastroenterology. 2012; 142(Suppl 1):S388. 29. Gisbert JP, Panes J. Loss of response and requirement of infliximab dose intensification in Crohn’s disease: a review. Am J Gastroenterol. 2009; 104:760–767. 30. Billioud V, Sandborn WJ, Peyrin-Biroulet L. Loss of response and need for adalimumab dose intensification in Crohn’s disease: a systematic review. Am J Gastroenterol. 2011;106:674–684.
www.ibdjournal.org |
1295
