REVIEW

Immunogenicity of anti-TNF biologic agents in the treatment of rheumatoidarthritisChi Chiu Moka, Wen Chan Tsaib, Der Yuan Chenc,d and James Cheng Chung Weie,f

aDepartment of Medicine, Tuen Mun Hospital, Hong Kong, China; bKaohsiung Medical University Hospital, Kaohsiung City, Taiwan;cDepartment of Medical Education and Research, Division of Allergy, Immunology and Rheumatology, Taichung Veterans General Hospital,Taichung, Taiwan; dFaculty of Medicine, National Yang Ming University, Taiwan and Institute of Biomedical Science, National Chung HsingUniversity, Taichung, Taiwan; eInstitute of Medicine, Chung Shan Medical University; Division of Allergy, Immunology and Rheumatology,Chung Shan Medical University Hospital, Taichung, Taiwan; fInstitute of Integrative Medicine, China Medical University, Taichung, Taiwan

ABSTRACTIntroduction: The use of biologic disease-modifying anti-rheumatic drugs (DMARDs), includingtherapeutic antibodies, antibody fragments and protein constructs that target key mediators inthe pathophysiology of rheumatoid arthritis (RA), has improved the chance of achieving lowdisease activity and clinical remission. However, individual patients respond differently to biologicDMARD therapy, particularly the tumor necrosis factor (TNF) inhibitors.Areas covered: While the variation of clinical response may be related to pharmacogenetic andother unknown factors, immunogenicity associated with some of these agents may contribute inpart to a lack of efficacy and immune-mediated side effects. Timely detection of immunogenicitymay avoid continued administration of ineffective treatment, and reduce unnecessary risks andcosts. Access to and appropriate implementation of clinically validated drug level assays isrequired.Expert opinion: There are currently no evidence-based recommendations to guide biologictherapy on the basis of drug level and immunogenicity testing but as more data becomeavailable and better tests are developed, a strategy of immunopharmacologic guidance toindividualize treatment of RA will emerge. The potential benefits of this approach must bebalanced against the costs of monitoring, and further research is required.

ARTICLE HISTORYReceived 15 June 2015Accepted 6 October 2015

KEYWORDSAntidrug antibodies; anti-TNF therapies;immunogenicity;rheumatoid arthritis

1. Introduction

Biologic disease-modifying anti-rheumatic drugs(DMARDs) represent a major breakthrough in the treat-ment of rheumatoid arthritis (RA). Since the late 1990s,the introduction of a number of agents that target andblock key protein messenger molecules (e.g. tumornecrosis factor-α [TNF-α], interleukin-6 [IL-6]) or cells(such as B-lymphocytes) involved in the pathogenesisof RA has made it possible to achieve very low diseaseactivity or even clinical remission, end points that havebeen adopted as realistic goals in current treatmentguidelines.[1–4]

The TNF inhibitors, a T-cell co-stimulation inhibitor(abatacept), a B-cell inhibitor (rituximab) and an IL-6antagonist (tocilizumab), are the currently available bio-logical agents for the treatment of RA.[3] Globally, theTNF inhibitors tend to be the first agents prescribedwhen biologic DMARDs are indicated in RA, due to thewealth of evidence, experience and long-term follow-up

data.[3,5] Today, five TNF inhibitors are available in themarket: infliximab, etanercept, adalimumab, golimumaband certolizumab pegol. In recent years, biosimilar ver-sions of the older branded biologic DMARDs have alsobecome available in some countries. Yisaipu, for exam-ple, is an etanercept biosimilar marketed in China,[6] andCT-P13, an infliximab biosimilar, has been marketed inSouth Korea.[7] Other etanercept and rituximab biosimi-lars are in clinical trials in other Asian countries.[8]

While there is no evidence from clinical trials thatany TNF inhibitor is more effective than any other, acertain proportion of patients do not respond ade-quately as judged from their clinical response. Thereasons for the refractoriness to the TNF inhibitorsremain to be elucidated and may be related to phar-macogenetic and many yet unknown factors. One ofthe potential reasons for the lack or loss of efficacy ofthe TNF inhibitors over time is the immunogenicityassociated with biologic DMARDs.[9–11] This article

CONTACT James Cheng-Chung Wei jccwei@gmail.com Institute of Medicine Chung Shan Medical University Hospital, No. 110, Sec. 1, Jianguo N.Rd., South District, Taichung City 40201, Taichung, TaiwanChi Chiu Mok and James Cheng Chung Wei contributed equally to the article.

EXPERT OPINION ON BIOLOGICAL THERAPY, 2016VOL. 16, NO. 2, 201–211http://dx.doi.org/10.1517/14712598.2016.1118457

© 2015 Taylor & Francis

Dow

nloa

ded

by [

Nat

iona

l Tai

wan

Uni

vers

ity]

at 0

0:29

14

May

201

6

reviews the current understanding of the immunogeni-city of biologic agents used in the treatment of RA, andthe implications for clinical practice and treatment out-comes in Asian countries.

2. Basic concepts in immunogenicity ofbiologic DMARDs

Immunogenicity is the ability of a substance to provokean immune response when recognized by the body asbeing foreign. Biologic therapies used in RA are engi-neered molecules such as therapeutic antibodies, anti-body fragments or protein constructs. Therapeuticproteins are a common cause of immune reactions;therefore, it is not surprising that biologic DMARDshave the potential to be immunogenic, and can resultin the formation of antibodies against the drug, knownas antidrug antibodies.[9]

The first therapeutic antibodies developed were ofmouse origin and resulted in the production of humanantimouse antibodies in 90% of treated patients, sub-stantially limiting their clinical use.[12] Newer treat-ments are therapeutic antibodies and antibodyfragments that are partly human derived – such asthe chimeric human–mouse monoclonal antibody(mAb) infliximab – or mAbs such as adalimumab andgolimumab, which are mostly human derived. Thesetherapies have lower immunogenicity than the origi-nal murine antibodies, but even fully human therapeu-tic antibodies can produce an immune response.[13]The mechanisms involved in immunogenicity againstbiologics are complex, and their manifestations andclinical implications vary among patients; effects canbe persistent or transient and are influenced by manyfactors other than the composition of the drugitself.[9]

3. Factors influencing the development ofimmunogenicity

The high variation in the prevalence of antidrug antibo-dies in RA patients using the TNF inhibitors reported indifferent studies suggests that patient factors may play animportant role in immunogenicity and its manifestations.Immunogenicity is also influenced by the underlying dis-ease, the treatment schedule ‒ drug dose, administrationroute, frequency, continuity ‒ the drug’s physicochemicalproperties (for example, purity, aggregation) and conco-mitantmedications, in particular where biologics are com-monly used in conjunction with traditional DMARDs(Table 1).[9,11,12,14–33] The observation of immunogeni-city may also be related to assaymethods (see Section 4.3)and the timing of serum sampling.[14,34,35]

The reported frequency of antidrug antibodies tomAbs varies across different diseases (Table 2),[15,32,36–54] although the wide range is also the resultof other factors described above, and the use of con-comitant immunomodulators.[30]

4. Immunogenicity of TNF inhibitors and itspotential clinical implications

4.1. Differences in the immunogenic regions ofmAbs and soluble fusion proteins

The most recent addition to the TNF inhibitor class iscertolizumab pegol, a PEGylated Fab fragment of ahumanized anti-TNF antibody.[5] Its lack of an Fc regionavoids potential Fc-mediated effects such as comple-ment- or antibody-dependent, cell-mediated cytotoxi-city; attachment of the PEG moiety to the Fab fragmentextends the plasma half-life to approximately 2 weeks.[55] Potential immunogenic sites on the TNF inhibitormolecules are summarized in Table 3.[56]

4.2. Potential impact of immunogenicity on thepharmacokinetics of TNF inhibitors

The neutralizing effect of antidrug antibodies on TNF inhi-bitors is evident from their inverse relationship with serumdrug levels, which was first demonstrated in RA patientsreceiving intravenous infliximab.[36] Development of anti-drug antibodies resulted in low serum drug levels and wasassociated with treatment failure, need for dose increases,immune-mediated adverse effects and cessation of ther-apy. Several other studies in patients treated with inflix-imab or adalimumab have also shown that thedevelopment of antidrug antibodies over time correlateswith the disappearance of drug from the blood.[35,56]

Article highlights

● Individual RA patients respond differently to biologic DMARDtherapy, particularly the TNF inhibitors. Lack or loss of efficacyand immune-mediated side effects may be partly attributableto the immunogenicity associated with some of these agents.

● Timely detection of immunogenicity may avoid continuedadministration of ineffective treatment, and reduce unnecessaryrisks and costs.

● There are currently no evidence-based recommendations toguide biologic therapy on the basis of drug level and immuno-genicity testing, but an immunopharmacologic approach toindividualizing treatment is likely to emerge in the future.

● The potential benefits of this approach need to be balancedagainst the costs of monitoring.

This box summarizes key points contained in the article

202 C. C. MOK ET AL.

Dow

nloa

ded

by [

Nat

iona

l Tai

wan

Uni

vers

ity]

at 0

0:29

14

May

201

6

Subsequent administration of a drug in the presenceof antidrug antibodies can result in the formation ofimmune complexes, which may affect the ability of thedrug to bind to its therapeutic target. These neutraliz-ing antidrug antibodies may also reduce therapeuticefficacy indirectly by reducing the release of the drugfrom the injection site into the circulation, therebycompromising bioavailability.[56,57] For biologics admi-nistered by subcutaneous injection, development of

antidrug antibodies near the injection site can contri-bute to local formation of immune complexes.[56]

The incidence of antidrug antibodies to the fusionprotein etanercept is much lower than that reportedwith the monoclonal anti-TNF antibodies (6% accordingto one study [19] in RA and 0% in other studies.[17,18])The occurrence of anti-etanercept antibodies does notappear to affect the clinical efficacy (non-neutralizing)of the TNF inhibitor.[19]

Table 1. Factors that may affect the development of immunogenicity to biologic DMARDs.Factor Evidence

Timing of serum sampling for antidrugantibody assay

● Among RA patients treated with adalimumab, 67% of those who develop antidrug antibodies do sowithin the first 28 weeks of treatment, meaning that 1 of 3 patients will develop antidrug antibodiesmuch later if therapy is continued [14]

● 2 out of 3 infliximab-treated RA patients who develop antidrug antibodies do so within the first22 weeks of treatment, although additional cases are detected up to week 46 [15]; 14% of patientswho develop antidrug antibodies do so >1 year after starting therapy [16]

Detection methods ● The use of different assays for antidrug antibodies contributes to variation in the frequency ofantibodies reported in different studies [9]

Drug structure ● Immunogenicity is markedly reduced with the more humanized mAbs compared to chimeric mAbs[12]

● The greatest structural difference among the TNF inhibitors is between the mAb-based drugs andetanercept, a soluble TNF receptor–Fc Ig fusion protein; although formation of antidrug antibodies isdescribed in the etanercept label, these antibodies are not neutralizing and do not affect clinicalefficacy [11,17–20]

Mode of administration Subcutaneous (SC) vs. intravenous (IV)● The SC route is potentially linked to increased risk of immunogenicity as a result of differences in

antigen presentation● Abatacept is the only biologic DMARD available for both SC and IV administration; immunogenicity in

RA patients is comparably low for both routes [21–23]Continuous vs. intermittent use● Intermittent use of biologics may induce a higher frequency of antidrug antibodies than continuous

therapy [11]● In RA patients, the risk of serious infusion reaction to infliximab (thought to be related to immune

complex formation between the drug and antidrug antibodies) is higher when patients are retreatedafter a prolonged interval off treatment [24]

● A non-significant increase in immunogenicity may occur when abatacept therapy is interrupted inpatients with RA, but the effect may be reversed when abatacept is reintroduced [21]

Genetic polymorphism ● The role of genetic factors in immunogenicity to biologic DMARDs in RA is unclear [9]● Genetic predisposition is suggested by the fact that patients who show an immune reaction to one TNF

inhibitor are more likely to also develop antidrug antibodies to another TNF inhibitor [25,26]● Formation of antibodies against adalimumab in RA patients has been associated with IL-10 gene

polymorphisms [27]Underlying disease ● Patients with certain autoimmune diseases (e.g. systemic lupus erythematosus [SLE]) are more prone

to develop antidrug antibodies. One study shows that 1 of 3 SLE patients develop antibodies to thechimeric mAb rituximab after administration for 1 year, which is associated with poor B-cell depletion[28]

Disease/ inflammatory activity ● Patients with high baseline disease activity of RA are more prone to the development of antidrugantibody after being treated with adalimumab. Higher incidence of anti-adalimumab antibody isobserved in patients with more longstanding and severe (often with erosion) disease at baseline, andhigher disease activity score [29]

Concomitant immunomodulators ● Co-administration of methotrexate or azathioprine reduces the immunogenicity of therapeutic anti-bodies in RA and other chronic inflammatory conditions, although the mechanism is unclear [30,31]

● In a systematic review, among patients with immune-mediated inflammatory diseases (RA, ankylosingspondylitis, psoriasis and inflammatory bowel disease), the use of methotrexate, azathioprine ormercaptopurine reduces the frequency of antidrug antibodies to TNF inhibitors by 47% and attenuatestheir impact on clinical response [20]

● In patients with Crohn’s disease treated with infliximab, concomitant azathioprine or methotrexatesignificantly reduces the incidence of antidrug antibodies and prolongs the duration of responsecompared with infliximab monotherapy; in the absence of antidrug antibodies, concomitantDMARDs do not influence the duration of response, indicating that the effect is attributable toimmunogenic modulation [32]

● In a 3-year longitudinal cohort of 272 RA patients treated with adalimumab, the development of anti-adalimumab antibody correlates inversely with the weekly dose of concomitant methotrexate [33]

DMARD: Disease-modifying anti-rheumatic drug; Fc: Fragment crystallizable region; Ig: Immunoglobulin; IL: Interleukin; mAb: Monoclonal antibody; RA:Rheumatoid arthritis; TNF: Tumor necrosis factor.

EXPERT OPINION ON BIOLOGICAL THERAPY 203

Dow

nloa

ded

by [

Nat

iona

l Tai

wan

Uni

vers

ity]

at 0

0:29

14

May

201

6

4.3. Immunogenicity testing

While immunogenicity may explain some of the differ-ences among the TNF inhibitors, the lack of standardizedassays has led to the great variation in the prevalence ofantidrug antibodies to these agents. The specificities andsensitivities of available assays differ, making comparisonsbetween studies difficult.[9] Standard enzyme-linkedimmunosorbent assay (ELISA) techniques are the leastspecific, and do not distinguish between functionallyactive and inactive antidrug antibodies.[9,56] Improvedmethods are two-site (bridging) assays and radioimmu-noassay (RIA)-based antigen binding tests.[58] However,the bridging assay does not detect certain types of anti-drug antibody that are more prominent after prolongeddrug exposure (IgG4).[56] A comparison of various assaymethods in RA patients treated with infliximab concludedthat RIA- and ELISA-based solid-phase cross-binding testswere inferior to fluid-phase RIA.[59]

The measurement of antibodies against antibodies isfurther complicated by the presence of the drug itself.Current antibody assays measure unbound antibodies,

not those bound in an immune complex; therefore,when both drug and antidrug antibody are present,formation of immune complexes between the tworesults in lower levels of free antidrug antibody (aswell as lower levels of unbound drug) and the testwill indicate a falsely low level. Conversely, in the pre-sence of antidrug antibodies, drug concentration mea-surements may be incorrect.[9] Interpretation ofantibody assessments should therefore make referenceto the dosing and timing of the drug, and to thecorresponding serum drug levels.[9] To minimize inter-ference with antidrug antibody analysis, blood samplesshould be collected immediately prior to drug adminis-tration, when drug levels are at their lowest.[60] Evenso, individual trough levels may vary. To overcome suchinteractions, new methods (reporter gene assays) arebeing explored that directly measure TNF neutralizationand its inhibition by antidrug antibodies.[61] Assaysusing liquid chromatography with tandem mass spec-trometry are also being developed (Wen-Chan Tsai,unpublished observations).

Table 2. Reported frequencies of antidrug antibodiesa to infliximab and adalimumab in various inflammatory diseases.Inflammatory disease Percentage of patients with antibodies to infliximab Percentage of patients with antibodies to adalimumab References

Rheumatoid arthritis 8–51 12–37 [15,36–42]Crohn’s disease 14–61 0.04b–17 [32,43–49]Juvenile idiopathic arthritis NA 16 [50]Ankylosing spondylitis 29 31 [51,52]Psoriatic arthritis NA 18 [53]Psoriasis vulgaris NA 45 [54]

Adapted with permission from [30]. Published by BioMed Central.NA: not applicable.aOverall frequency in each study was taken when developing this table; in some studies, frequency was shown to vary with dose or when infliximab/adalimumab was given in combination with another drug.

bThe low rate of anti-adalimumab antibodies reported by Hanauer et al. [49] may have been related to the relatively short duration of the study (4 weeks).

Table 3. Summary of immunogenic sites of TNF inhibitors for RA.

Drug(and year offirst approvalfor use in RA) Structure

Mode ofadministration

Potential immunogenic sites against which antidrug antibodies may be produced

Mouseepitopeon drug

FR

Mouseepitopeon drugCDR

Human idiotopesin or on drugbinding site to

TNF-α

Humanallotopes onconstant

regions of IgGNonhuman

glycosylations

Neoepitopeson drug

aggregates

Splice-regionpeptide

Etanercept(1998)

TNF receptor–Fcfusion protein

SC ✓ ✓ ✓ ✓

Infliximab(1999)

Chimeric human–mouse mAb

IV ✓ ✓ ✓ ✓ ✓

Adalimumab(2002)

Human mAb SC ✓ ✓ ✓ ✓

Golimumab(2009)

Human mAb SC ✓ ✓ ✓ ✓

Certolizumabpegol(2009)

PEGylatedrecombinanthumanized Fabfragment of anti-TNFmAb

SC ✓ ✓ ✓

Adapted with permission from [56].CDR: Complementarity-determining variable region; Fab: Fragment antigen binding; Fc: Fragment crystallizable region; FR: Framework region; IgG:Immunoglobulin; IV: Intravenous; mAb: Monoclonal antibody; PEG: Polyethylene glycol; RA: Rheumatoid arthritis; SC: Subcutaneous; TNF: Tumor necrosisfactor.

Shaded rows indicate mAbs.

204 C. C. MOK ET AL.

Dow

nloa

ded

by [

Nat

iona

l Tai

wan

Uni

vers

ity]

at 0

0:29

14

May

201

6

5. Potential clinical consequences ofimmunogenicity

5.1. Clinical efficacy

Evidence of the consequences of immune responses tobiologic DMARDs has grown substantially during thelast decade,[34] largely based on experience with inflix-imab [15,16,36,40] and adalimumab.[14,25,35,37] Thesestudies consistently show that, in patients with RA,immunogenicity is strongly linked to sub-therapeuticserum drug levels and a compromise of clinicalresponse. A 3-year longitudinal study of 407 RA patientsshowed that etanercept and adalimumab are similar ininducing a good long-term clinical outcome, but foradalimumab this is strongly dependent on the presenceof antidrug antibodies.[62] Patients with anti-adalimu-mab antibodies had the lowest rate of sustained mini-mal disease activity compared to those withoutantidrug antibodies or treated with etanercept.[62]

Another study of 147 RA patients treated with inflix-imab for at least 6 months showed an interesting find-ing. Among those who achieved low disease activity,one-third of patients had anti-infliximab antibodies,with half of these patients having these antibodiesevery visit for more than 1 year[63] Thus, monitoringof infliximab trough levels and the presence of anti-infliximab antibodies can assist in the clinical decisionof down-titration and discontinuation of infliximab inpatients with stable low disease activity state so thatdrug cost can be minimized.[64]

Data on the long-term consequences of immuno-genicity of the anti-TNF biologics in RA are availablefrom open-label extensions of randomized clinical trialsand observational registries. For example, in a 3-yearfollow-up of 272 RA patients treated with adalimumab,development of antidrug antibodies was associatedwith lower adalimumab concentration and lower like-lihood of achieving minimal disease activity or clinicalremission.[14] Information regarding immunogenicityof the newer TNF inhibitors such as golimumab andcertolizumab is less well reported in the literature.

As it takes time to produce an antibody response toan exogenous protein, drug immunogenicity is not alikely explanation for most primary response failures tothe TNF inhibitors; however, it is a major contributorto secondary response failure, which affects up to 50%of patients with good primary responses to TNF inhibi-tors.[56]

Limited information on immunogenicity is availablein clinical trials of TNF inhibitors originating from Asiancountries. In the multicenter study of adalimumab

monotherapy in RA patients in Japan (CHANGE),approximately one-third of patients treated with adali-mumab expressed antidrug antibodies at week 24 andthis was associated with poorer clinical response(defined by American College of Rheumatology criteriafor 20% improvement) when compared to those with-out antidrug antibodies.[41] The efficacy and safety ofadalimumab monotherapy in Japanese patients wassimilar to that observed in a similar study at compar-able doses in Western patients,[42] except that therewas a higher frequency of injection-site reactions in theJapanese study (25% vs. 10%). However, for unknownreasons, mean steady-state serum concentrations ofadalimumab in Japanese patients were lower than inthe study of Western patients, and the incidence ofanti-adalimumab antibodies was higher.[41]

A recent study examined the effect of antidrug anti-bodies on the clinical efficacy and withdrawal rate ofinfliximab, adalimumab and etanercept in 58 Chinesepatients with rheumatic diseases.[65] The proportion ofpatients who developed antidrug antibodies was 50, 31and 0%, respectively. Those with antidrug antibodies hadsignificantly lower serum drug levels and a significantlyhigher drug withdrawal rate due to lack of efficacy. Inpatients with RA or psoriatic arthritis, non-response wassignificantly more frequent in those with antidrug anti-bodies than in those without (54% vs. 13%; p = 0.01).Treatment significantly improved disease activity score inpatients with spondyloarthritis or ankylosing spondylitiswho were antidrug antibody-negative, but not in thosewho were antidrug antibody-positive.[65]

Epidemiologic studies of RA in Asia are scarce,[66]but data on the use of biologics are beginning toemerge from patient registries and national databases.A registry-based study for RA patients in Taiwanrevealed that anti-adalimumab antibodies weredetected in 27.8 and 36.1% of adalimumab-treatedpatients after 12-month therapy using ELISA and radio-immunoassay, respectively, but were not detected inany of the etanercept-treated patients. The presenceof anti-adalimumab antibodies was associated withlower drug trough levels and lower therapeuticresponse compared to those without anti-adalimumabantibodies.[35] A registry-based study in Taiwandemonstrated that high-titer (>100 AU) and low-titer(12 – 100 AU) anti-adalimumab antibodies can bedetected in 20 and 63%, respectively, of ankylosingspondylitis patients after 44 weeks of treatment. Theseantibodies were associated with decreased serum druglevels and reduced clinical efficacy.[67] An observa-tional cohort study of 115 patients with ankylosing

EXPERT OPINION ON BIOLOGICAL THERAPY 205

Dow

nloa

ded

by [

Nat

iona

l Tai

wan

Uni

vers

ity]

at 0

0:29

14

May

201

6

spondylitis treated with adalimumab in the Netherlandsand Taiwan found that 27% of patients had detectableanti-adalimumab antibodies at week 24. Mean adalimu-mab levels were significantly higher in patients withoutantibodies than in those with antibodies (12.7 vs.1.2 mg/l; p < 0.001) and there was a significant relation-ship between adalimumab levels and clinical response(p = 0.02).[68]

5.2. Infusion reaction

Potential adverse effects mediated by antidrug anti-body immune complexes include infusion-relatedhypersensitivity reactions, which have been observedin up to half of RA patients expressing anti-infliximabantibodies [24,30,69] and 20% of infliximab-treatedpatients across all clinical trials.[70] Such reactionshave been linked to the formation of infliximab–anti-infliximab complexes,[71] with higher antibody con-centrations associated with a higher risk of infusionreactions.[43] The majority of infusion reactions aremild to moderate, occur during or within 2 hours ofthe infusion, and can be managed by slowing or inter-rupting the infusion and administering antihistaminesand/or corticosteroids. However, more severe hyper-sensitivity and anaphylactic reactions have beenreported, including laryngeal/pharyngeal edema andsevere bronchospasm, and require appropriate emer-gency treatment. The risk of serious infusion reactionsmay be higher when patients are retreated after aprolonged interval off treatment; moreover, the reac-tion may be a delayed one, commonly occurring afterinitiation of the second infusion.[24] Serious infusionreactions have not been reported with etanercept,although mild to moderate injection-site reactions(erythema, itching, pain, swelling, bleeding and bruis-ing) lasting 3 – 5 days are the most common adverseevent; these generally occur in the first month andthen become less frequent.[72] Whether the etiologyof this reaction is irritative or immune-mediated is notfully understood.[73]

5.3. Other adverse consequences

Serum sickness and Arthus reactions are local immunecomplex-mediated type III hypersensitivity reactionsthat are occasionally associated with immunogenicityof anti-TNF mAb therapy.[74] With infliximab, serumsickness-like reactions have been observed as early asthe second dose, and also when therapy is restartedafter an extended period off treatment. These reactionsare associated with a marked increase in antibodies toinfliximab, and are characterized by fever, rash,

headache, sore throat, myalgias, polyarthralgias, handand facial edema, and dysphagia.[70]

Immunogenic reactions against biologics have alsobeen associated with thromboembolic events. A cohortstudy of adalimumab-treated RA patients reported thatthromboembolic events were more commonlyobserved in those who tested positive for the anti-adalimumab antibody.[75] The authors suggested thatdrug–antidrug immune complexes might induce aggre-gation and procoagulant microparticle release.However, this observation has to be confirmed byother groups and future studies.

6. ‘Cross-reactivity’ of the antidrug antibodies

A cohort study of 235 patients with RA treated withadalimumab reported that 20% of patients developedanti-adalimumab antibodies.[25] Fifty-two (22%)patients in this cohort were switchers from infliximabto adalimumab. Patients positive for anti-adalimumabantibodies had significantly lower reduction of diseaseactivity score in 28 joints (DAS28) at week 28 comparedto those without anti-adalimumab antibodies. It is alsointeresting that the incidence of anti-adalimumab anti-bodies was significantly higher in those switchers frominfliximab who had developed anti-infliximab antibo-dies than in those who were TNF inhibitor-naïve (33%vs. 18%; p = 0.04).[25] However, the improvement inDAS28 score was significantly lower in switchers with-out the anti-infliximab antibodies than in patients naïveto the TNF inhibitors. While further data are needed toconfirm this observation, this might suggest that, whentreatment failure to an anti-TNF agent is not because ofimmunogenicity, the chance of a clinical response to asecond anti-TNF agent is lower because these patientsare likely to be refractory to the drug mechanism of TNFblockade.

7. Retention rate of the TNF inhibitors andswitching of anti-TNF therapy

While short-term data suggest that the survival rates ofthe anti-TNF biologics may be similar, long-term obser-vational data show that etanercept exhibits the bestlong-term drug survival (60% at 4 years).[76] In theItalian GISEA registry, the 4-year global drug survivalof adalimumab, etanercept and infliximab was lowerthan 50%, but etanercept had the best retention rate.[77] Drug survival was also highest for etanercept in theDanish DANBIO registry.[78] Data on registries are nowavailable in Asia. In the biologic registry maintained bythe Hong Kong Society of Rheumatology, it wasreported that the 5-year retention rate of the anti-TNF

206 C. C. MOK ET AL.

Dow

nloa

ded

by [

Nat

iona

l Tai

wan

Uni

vers

ity]

at 0

0:29

14

May

201

6

agents for the treatment of rheumatic diseases waslower with infliximab than with adalimumab and eta-nercept.[79]

When anti-TNF therapy fails, options includeintensifying therapy, switching to another anti-TNFtherapy or switching to another class of drug; how-ever, there is no universally accepted approach andwithout knowing the precise reason for treatmentfailure it is effectively a trial-and-error strategy.[56]With infliximab, some patients benefit from raisingthe dose from 3 mg/kg every 8 weeks to 10 mg/kgas often as every 4 weeks when the clinical responsewanes over time (secondary treatment failure),[41]because the incidence of anti-infliximab antibodieshas been shown to correlate inversely with thedosage of infliximab being administered (7, 21 and53%, respectively, in patients treated with infliximab10 mg/kg, 3 mg/kg or 1 mg/kg [39,80]). However, ahigher dose of the anti-TNF monoclonals to over-come immunogenicity will result in higher treatmentcost incurred.[36,37,81] Conversely, patients whocontinue on TNF inhibitor therapy during clinicalremission may risk developing antidrug antibodiesthat compromise the efficacy of subsequent coursesof treatment. Identification of these patients willallow withdrawal of therapy that would help toreduce drug cost.[64]

Observational data from several national registriesand open-label studies suggest that switching TNFinhibitors is a useful strategy in RA patients who failtreatment with a first TNF inhibitor.[11,25,82–84]However, no immunogenicity data are available inthese studies; therefore, it is difficult to interpretwhether treatment failures were attributable to devel-opment of antidrug antibodies and the implicationsfor switching therapy.

A recent study of 292 consecutive RA patients trea-ted with etanercept showed that among those whowere switched from infliximab or adalimumab (30%),the presence of antidrug antibodies to the first TNFinhibitor was associated with a better clinical responsein terms of improvement in DAS28 score at week 28after etanercept therapy.[85] Thus, determining immu-nogenicity can help to identify patient subsets in whomswitching to another anti-TNF agent could bebeneficial.

8. Conclusions

The immunogenicity of anti-TNF therapies, even thosedescribed as fully human, influences their efficacy andsafety in patients with RA. Currently, the issue of immu-nogenicity of the biological DMARDs and the

monitoring of drug levels is not yet addressed in theupdated 2013 EULAR recommendation.[3] However,regular monitoring of immunopharmacologic statususing robust and clinically validated tests may facilitatea better understanding of the treatment response andenable individualization of the most appropriate ther-apy. Further research is necessary to better establish therelationship between drug levels, antidrug antibodiesand clinical efficacy. When clinical efficacy of the biolo-gical DMARDs decreases with time or adverse drugreactions emerge, measuring drug level and/or antidrugantibody levels could help to guide dosage adjust-ments and drug switching. When more data becomeavailable and better tests are developed, a strategy ofimmunopharmacologic guidance may become a stan-dard of care in the future and help improve the cost-effectiveness of the anti-TNF biologics in the treatmentof rheumatoid arthritis.

9. Expert opinion

The high cost of most biologic DMARDs precludestheir widespread use, particularly in developingregions of Asia. Therefore, it is prudent to recognizeintolerance or lack of response to TNF inhibitor ther-apy early. Given the high individual variability in thepharmacokinetics of the anti-TNF agents, and thedifficulty of predicting outcomes by clinical judgmentalone, the rationale for therapeutic drug level mon-itoring of biologics and assessment of antidrug anti-body responses is strengthening. However, the valueof immunopharmacologic guidance in RA is a topic ofongoing debate. To date, there are no evidence-based recommendations to guide selection or switch-ing of biologic therapies based on drug level andimmunogenicity testing after failure of anti-TNF ther-apy, and the benefits of adopting such an approachwould need to be balanced against the availability ofreliable and reproducible tests and the costs ofmonitoring.

In clinical practice, it is likely that plasma drug levelswill be useful in the monitoring of anti-TNF therapy.Sub-therapeutic plasma drug levels may be due to poortreatment adherence, pharmacogenetic reasons or thedevelopment of neutralizing antibody against the anti-TNF agent. If further testing confirms the presence ofantidrug antibodies, increasing the dose may improveresponse; however, the practice of intensifying therapyto restore lost efficacy is not always successful and thecost may be prohibitive. Furthermore, immunogenicitymay change over time and patients may either becometolerant to treatment or continue to produce furtherantibodies.

EXPERT OPINION ON BIOLOGICAL THERAPY 207

Dow

nloa

ded

by [

Nat

iona

l Tai

wan

Uni

vers

ity]

at 0

0:29

14

May

201

6

Therapeutic algorithms based on expert opinion sug-gest utilizing pharmacokinetic and immunogenic test-ing to guide anti-TNF treatment.[10,11,20,35,56] Wepropose a practical clinical algorithm for the monitoringof drug level and immunogenicity of the anti-TNFagents (Figure 1), based on our own opinions. As anti-drug antibody tests can produce false-negative results ifconducted too early after initiation of the TNF inhibi-tors, it is better to measure drug level and antidrugantibody level at baseline, and ideally every 6 months,or whenever clinical efficacy wanes.

Financial and competing interests disclosure

Medical writing support was provided by Susanne Gilbert atACUMED, an Ashfield Company, and was funded by Pfizer Inc.WC Tsai has received consultation fees from Abbvie, Chugai,Eisai, Janssen, Novartis and Pfizer. JCC Wei has receivedresearch grants or consultation fees from Abbvie, Bristol-Myers Squibb, Celgene, Chugai, Eisai, Janssen, Novartis,Pfizer, Sanofi-Aventis, TSH Taiwan and UCB Pharma. Theauthors have no other relevant affiliations or financial involve-ment with any organization or entity with a financial interestin or financial conflict with the subject matter or materialsdiscussed in the manuscript. This includes employment,

consultancies, honoraria, stock ownership or options, experttestimony, grants or patents received or pending, or royalties.

Bibliography

Papers of special note have been highlighted as either ofinterest (•) or of considerable interest (••) to readers.1. National Collaborating Centre for Chronic Conditions.

Rheumatoid arthritis: national clinical guideline for man-agement and treatment in adults. London: Royal Collegeof Physicians; 2009.

2. Smolen JS, Aletaha D, Bijlsma JW, et al. Treating rheuma-toid arthritis to target: recommendations of an interna-tional task force. Ann Rheum Dis. 2010;69:631–637.

3. Smolen JS, Landewé R, Breedveld FC, et al. EULAR recom-mendations for the management of rheumatoid arthritiswith synthetic and biological disease-modifying anti-rheumatic drugs: 2013 update. Ann Rheum Dis.2014;73:492–509.

4. Singh JA, Furst DE, Bharat A, et al. Update of the 2008American College of Rheumatology recommendations forthe use of disease-modifying anti-rheumatic drugs andbiologic agents in the treatment of rheumatoid arthritis.Arthritis Care Res (Hoboken). 2012;2012(64):625–639.

5. Jung YO, Kim HA. Recent paradigm shifts in the diagnosisand treatment of rheumatoid arthritis. Korean J InternMed. 2012;27:378–387.

Figure 1. Proposed treatment algorithm for RA patients who stop responding to treatment with TNF inhibitor therapy.ADAb: Antidrug antibodies; mAb: Monoclonal antibody; RA: Rheumatoid arthritis; TNF: Tumor necrosis factor.

208 C. C. MOK ET AL.

Dow

nloa

ded

by [

Nat

iona

l Tai

wan

Uni

vers

ity]

at 0

0:29

14

May

201

6

6. Wu B, Song Y, Leng L, et al. Treatment of moderaterheumatoid arthritis with different strategies in ahealth resource-limited setting: a cost-effectivenessanalysis in the era of biosimilars. Clin Exp Rheumatol.2015;33:20–26.

7. Yoo DH, Hrycaj P, Miranda P, et al. A randomised, double-blind, parallel-group study to demonstrate equivalence inefficacy and safety of CT-P13 compared with innovatorinfliximab when coadministered with methotrexate inpatients with active rheumatoid arthritis: the PLANETRAstudy. Ann Rheum Dis. 2013;72:1613–1620.

8. Dörner T, Strand V, Castañeda-Hernández G, et al. The roleof biosimilars in the treatment of rheumatic diseases. AnnRheum Dis. 2013;72:322–328.

9. Wolbink GJ, Aarden LA, Dijkmans BA. Dealing with immu-nogenicity of biologicals: assessment and clinical rele-vance. Curr Opin Rheumatol. 2009;21:211–215.

10. Bendtzen K. Is there a need for immunopharmacologicguidance of anti-tumor necrosis factor therapies?Arthritis Rheum. 2011;63:867–870.

11. Vincent FB, Morand EF, Murphy K, et al. Anti-drug antibodies(ADAb) to tumour necrosis factor (TNF)-specific neutralizingagents in chronic inflammatory diseases: a real issue, a clin-ical perspective. Ann Rheum Dis. 2013;72:165–178.

12. Hwang WY, Foote J. Immunogenicity of engineered anti-bodies. Methods. 2005;36:3–10.

13. Clarck M. Antibody humanization: a case of the ‘Emperor’snew clothes’? Immunol Today. 2000;8:397–402.

14. Bartelds GM, Krieckaert CL, Nurmohamed MT, et al.Development of anti-drug antibodies against adalimumaband association with disease activity and treatment failureduring long-term follow-up. JAMA. 2011;305:1460–1468.

•• The first large prospective study on the effect of anti-drug antibody on the clinical efficacy of the anti-TNFmonoclonal.

15. Wolbink GJ, Vis M, Lems W, et al. Development of antiin-fliximab antibodies and relationship to clinical response inpatients with rheumatoid arthritis. Arthritis Rheum.2006;54:711–715.

16. Pascual-Salcedo D, Plasencia C, Ramiro S, et al. Influenceof immunogenicity on the efficacy of long-term treatmentwith infliximab in rheumatoid arthritis. Rheumatology.2011;50:1445–1452.

17. Daïen CI, Daïen V, Parussini E, et al. Etanercept concentra-tion in patients with rheumatoid arthritis and its potentialinfluence on treatment decisions: a pilot study. JRheumatol. 2012;39:1533–1538.

• A pilot study on the effect of drug level (etanercept)on clinical efficacy when used in patients with rheu-matoid arthritis.

18. de Vries MK, van der Horst-Bruinsma IE, Nurmohamed MT,et al. Immunogenicity does not influence treatment withetanercept in patients with ankylosing spondylitis. AnnRheum Dis. 2009;68:531–535.

19. Dore RK, Mathews S, Schechtman J, et al. The immuno-genicity, safety, and efficacy of etanercept liquid adminis-tered once weekly in patients with rheumatoid arthritis.Clin Exp Rheumatol. 2007;25:40–46.

20. Garcês S, Demengeot J, Benito-Garcia E. The immunogeni-city of anti-TNF therapy in immune-mediated inflamma-tory diseases: a systematic review of the literature with ameta-analysis. Ann Rheum Dis. 2013;72:1947–1955.

• A meta-analysis of 17 studies on the effect of antidrugantibodies on drug response in various immune-mediated inflammatory diseases.

21. Kaine J, Gladstein G, Strusberg I, et al. Evaluation of aba-tacept administered subcutaneously in adults with activerheumatoid arthritis: impact of withdrawal and reintroduc-tion on immunogenicity, efficacy and safety (Phase IIIbALLOW study). Ann Rheum Dis. 2012;71:38–44.

22. Schiff M. Subcutaneous abatacept for the treatment ofrheumatoid arthritis. Rheumatology (Oxford).2013;52:986–997.

23. Amano K, Matsubara T, Inoue H, et al. Long-term safetyand efficacy of treatment with subcutaneous abatacept inJapanese patients with RA who were MTX inadequateresponders – 76-week results. Ann Rheum Dis. 2013;72(Suppl. 3):458.

24. Toki H, Momohara S, Tsukahara S, et al. Infusion reactionto infliximab in a patient with rheumatoid arthritis afterdiscontinuation over 1 year and readministration. JRheumatol. 2008;35:1896–1897.

25. Bartelds GM, Wijbrandts CA, Nurmohamed MT, et al. Anti-infliximab and anti-adalimumab antibodies in relation toresponse to adalimumab in infliximab switchers and anti-tumour necrosis factor naive patients: a cohort study. AnnRheum Dis. 2010;69:817–821.

• Preliminary observation that the chance of antibodiesto one anti-TNF agent is higher in patients who havehistory of antibodies against other anti-TNF agents.

26. Schellekens H. The immunogenicity of therapeutic pro-teins. Discov Med. 2010;9:560–564.

27. Bartelds GM, Wijbrandts CA, Nurmohamed MT, et al. Anti-adalimumab antibodies in rheumatoid arthritis patientsare associated with interleukin-10 gene polymorphisms.Arthritis Rheum. 2009;60:2541–2542.

28. Albert D, Dunham J, Khan S, et al. Variability in the biolo-gical response to anti-CD20 B cell depletion in systemiclupus erythematosus. Ann Rheum Dis. 2008;67:1724–1731.

29. Krieckaert C, Rispens T, Wolbink G. Immunogenicity ofbiological therapeutics: from assay to patient. Curr OpinRheumatol. 2012;24:306–311.

30. Krieckaert CLM, Bartelds GM, Lems WF, et al. The effect ofimmunomodulators on the immunogenicity of TNF-block-ing therapeutic monoclonal antibodies: a review. ArthritisRes Ther. 2010;12:217.

31. Jani M, Barton A, Warren RB, et al. The role of DMARDsin reducing the immunogenicity of TNF inhibitors inchronic inflammatory diseases. Rheumatology (Oxford).2014;53:213–222.

32. Vermeire S, Noman M, Van Assche G, et al. Effectiveness ofconcomitant immunosuppressive therapy in suppressingthe formation of antibodies to infliximab in Crohn’s dis-ease. Gut. 2007;56:1226–1231.

33. Krieckaert CL, Nurmohamed MT, Wolbink GJ.Methotrexate reduces immunogenicity in adalimumabtreated rheumatoid arthritis patients in a dose dependentmanner. Ann Rheum Dis. 2012;71:1914–1915.

34. Krieckaert CLM, Lems WF. Are we ready for therapeuticdrug monitoring of biologic therapeutics? Arthritis ResTher. 2011;13:120.

35. Chen DY, Chen YM, Tsai WC, et al. Significant associationsof antidrug antibody levels with serum drug trough levelsand clinical effectiveness of adalimumab and etanercept

EXPERT OPINION ON BIOLOGICAL THERAPY 209

Dow

nloa

ded

by [

Nat

iona

l Tai

wan

Uni

vers

ity]

at 0

0:29

14

May

201

6

treatment in rheumatoid arthritis. Ann Rheum Dis.2015;74:e16.

36. Bendtzen K, Geborek P, Svenson M, et al. Individualizedmonitoring of drug bioavailability and immunogenicity inrheumatoid arthritis patients treated with the tumornecrosis factor inhibitor infliximab. Arthritis Rheum.2006;54:3782–3789.

37. Bartelds GM, Wijbrandts CA, Nurmohamed MT, et al.Clinical response to adalimumab: relationship to antiadali-mumab antibodies and serum adalimumab concentrationsin rheumatoid arthritis. Ann Rheum Dis. 2007;66:921–926.

38. Lipsky PE, van der Heijde DM, St Clair EW, et al. On behalfof the Anti-Tumor Necrosis Factor Trial in RheumatoidArthritis with Concomitant Therapy Study Group.Infliximab and methotrexate in the treatment of rheuma-toid arthritis. Anti-tumor necrosis factor trial in rheuma-toid arthritis with concomitant therapy study group. NEngl J Med. 2000;343:1594–1602.

39. Maini RN, Breedveld FC, Kalden JR, et al. Therapeuticefficacy of multiple intravenous infusions of anti-tumornecrosis factor alpha monoclonal antibody combinedwith low-dose weekly methotrexate in rheumatoid arthri-tis. Arthritis Rheum. 1998;41:1552–1563.

40. Radstake TR, Svenson M, Eijsbouts AM, et al. Formation ofantibodies against infliximab and adalimumab stronglycorrelates with functional drug levels and clinicalresponses in rheumatoid arthritis. Ann Rheum Dis.2009;68:1739–1745.

41. Miyasaka N. on behalf of the CHANGE Study Investigators.Clinical investigation in highly disease-affected rheuma-toid arthritis patients in Japan with adalimumab applyingstandard and general evaluation: the CHANGE study. ModRheumatol. 2008;18:252–262.

42. Van de Putte LBA, Atkins C, Malaise M, et al. Efficacy andsafety of adalimumab as monotherapy in patients withrheumatoid arthritis for whom previous disease modifyinganti-rheumatic drug treatment has failed. Ann Rheum Dis.2004;63:508–516.

43. Baert F, Noman M, Vermeire S, et al. Influence of immu-nogenicity on the long-term efficacy of infliximab inCrohn’s disease. N Engl J Med. 2003;348:601–608.

44. Hanauer SB, Feagan BG, Lichtenstein GR, et al. on behalfof the ACCENT I Study Group. Maintenance infliximab forCrohn’s disease: the ACCENT I randomised trial. Lancet.2002;359:1541–1549.

45. Sands BE, Anderson FH, Bernstein CN, et al. Infliximabmaintenance therapy for fistulizing Crohn’s disease. NEngl J Med. 2004;350:876–885.

46. Karmiris K, Paintaud G, Noman M, et al. Influence oftrough serum levels and immunogenicity on long-termoutcome of adalimumab therapy in Crohn’s disease.Gastroenterology. 2009;137:1628–1640.

47. Sandborn WJ, Hanauer SB, Rutgeerts P, et al. Adalimumabfor maintenance treatment of Crohn’s disease: results ofthe CLASSIC II trial. Gut. 2007;56:1232–1239.

48. West RL, Zelinkova Z, Wolbink GJ, et al. Immunogenicitynegatively influences the outcome of adalimumab treat-ment in Crohn’s disease. Aliment Pharmacol Ther.2008;28:1122–1126.

49. 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.

50. Lovell DJ, Ruperto N, Goodman S, et al. On behalf of thePediatric Rheumatology Collaborative Study Group and thePediatric Rheumatology International Trials Organisation.Adalimumab with or without methotrexate in juvenilerheumatoid arthritis. N Engl J Med. 2008;359:810–820.

51. De Vries MK, Wolbink GJ, Stapel SO, et al. Decreasedclinical response to infliximab in ankylosing spondylitis iscorrelated with anti-infliximab formation. Ann Rheum Dis.2007;66:1252–1254.

52. De Vries MK, Brouwer E, van der Horst-Bruinsma IE, et al.Decreased clinical response to adalimumab in ankylosingspondylitis is associated with antibody formation. AnnRheum Dis. 2009;68:1787–1788.

53. Van Kuijk AW, de Groot M, Stapel SO, et al. Relationshipbetween the clinical response to adalimumab treatmentand serum levels of adalimumab and anti-adalimumabantibodies in patients with psoriatic arthritis. Ann RheumDis. 2010;69:624–625.

54. Lecluse LL, Driessen RJ, Spuls PI, et al. Extent and clinicalconsequences of antibody formation against adalimumabin patients with plaque psoriasis. Arch Dermatol.2010;146:127–132.

55. Smolen J, Landewé RB, Mease P, et al. Efficacy and safetyof certolizumab pegol plus methotrexate in active rheu-matoid arthritis: the RAPID 2 study. A randomized con-trolled trial. Ann Rheum Dis. 2009;68:797–804.

56. Bendtzen K. Anti-TNF-α biotherapies: perspectives for evi-dence-based personalized medicine. Immunotherapy.2012;4:1167–1179.

57. Tracey D, Klareskog L, Sasso EH, et al. Tumor necrosisfactor antagonist mechanisms of action: a comprehensivereview. Pharmacol Ther. 2008;117:244–279.

58. Aarden L, Ruuls SR, Wolbink G. Immunogenicity of anti-tumor necrosis factor antibodies – toward improvedmethods of anti-antibody measurement. Curr OpinImmunol. 2008;20:431–435.

59. Svenson M, Geborek P, Saxne T, et al. Monitoring patientstreated with anti-TNF-biopharmaceuticals: assessing seruminfliximab and anti-infliximab antibodies. Rheumatology.2007;46:1828–1834.

60. Hart MH, De Vrieze H, Wouters D, et al. Differential effectof drug interference in immunogenicity assays. J ImmunolMethods. 2011;372:196–203.

• A study which shows that different assay methodsmay affect the incidence and levels of antibodies tothe anti-TNF monoclonals.

61. Lallemand C, Kavrochorianou A, Steenholdt C, et al.Reporter gene assay for the quantification of the activityand neutralizing antibody response to TNF- antagonists. JImmunol Meth. 2011;373:229–239.

62. Krieckaert CL, Jamnitski A, Nurmohamed MT, et al.Comparison of long-term clinical outcome with etaner-cept treatment and adalimumab treatment of rheumatoidarthritis with respect to immunogenicity. Arthritis Rheum.2012;64:3850–3855.

63. van der Maas A, van den Bemt BJF, Wolbink G, et al. Lowinfliximab serum trough levels and anti-infliximab antibo-dies are prevalent in rheumatoid arthritis patients treatedwith infliximab in daily clinical practice: results of an

210 C. C. MOK ET AL.

Dow

nloa

ded

by [

Nat

iona

l Tai

wan

Uni

vers

ity]

at 0

0:29

14

May

201

6

observational cohort study. BMC Musculoskelet Disord.2012;13:184.

64. van den Bemt BJF, den Broeder AA, Wolbink GJ, et al. Anti-infliximab antibodies are already detectable in mostpatients with rheumatoid arthritis halfway through aninfusion cycle: an open-label pharmacokinetic cohortstudy. BMC Musculoskelet Disord. 2011;12:12.

65. Mok CC, van der Kleij D, Wolbink GJ. Drug levels, anti-drugantibodies, and clinical efficacy of the anti-TNFα biologicsin rheumatic diseases. Clin Rheumatol. 2013;32:1429–1435.

66. Lai CH, Lai MS, Lai KL, et al. Nationwide population-basedepidemiologic study of rheumatoid arthritis in Taiwan.Clin Exp Rheumatol. 2012;30:358–363.

67. Yeo KJ, Leong PY, Chen CY, et al. Immunogenicity ofadalimumab in ankylosing spondylitis in Taiwan: A regis-try-based cohort study. [Submitted for publication]

68. Kneepkens EL, Wei JC, Nurmohamed MT, et al.Immunogenicity, adalimumab levels and clinicalresponse in ankylosing spondylitis patients during24 weeks of follow-up. Ann Rheum Dis. 2013;72(Suppl.s3):518.

69. Ducourau E, Mulleman D, Paintaud G, et al. Antibodiestoward infliximab are associated with low infliximab con-centration at treatment initiation and poor infliximabmaintenance in rheumatic diseases. Arthritis Res Ther.2011;13:R105.

70. REMICADE® (infliximab) Prescribing Information. JanssenBiotech, Inc. 2013. Available from: http://www.remicade.com/shared/product/remicade/prescribing-information.pdf

71. Van der Laken CJ, Voskuyl AE, Roos JC, et al. Imaging andserum analysis of immune complex formation of radiola-belled infliximab and anti-infliximab in responders andnon-responders to therapy for rheumatoid arthritis. AnnRheum Dis. 2007;66:253–256.

72. Enbrel® (etanercept) Prescribing Information 2013.Amgen. September 2013. Available from: http://pi.amgen.com/united_states/enbrel/derm/enbrel_pi.pdf

73. Batycka-Baran A, Flaig M, Molin S, et al. Etanercept-inducedinjection site reactions: potential pathomechanismsand clinical assessment. Expert Opin Drug Saf 77.2012;11:911–921.

74. Descotes J, Gouraud A. Clinical immunotoxicity of thera-peutic proteins. Expert Opin Drug Metab Toxicol.2008;4:1537–1549.

75. Korswagen LA, Bartelds GM, Krieckaert CL, et al. Venousand arterial thromboembolic events in adalimumab-treated patients with antiadalimumab antibodies: acase series and cohort study. Arthritis Rheum.2011;63:877–883.

76. Markatseli TE, Alamanos Y, Saougou I, et al. Survival ofTNF-alpha antagonists in rheumatoid arthritis: a long-termstudy. Clin Exp Rheumatol. 2012;30:31–38.

77. Iannone F, Gremese E, Atzeni F, et al. Long term retentionof tumor necrosis factor-α inhibitor therapy in a largeItalian cohort of patients with rheumatoid arthritis fromthe GISEA registry: an appraisal of predictors. J Rheumatol.2012;39:1179–1184.

78. Hetland ML, Christensen IJ, Tarp U, et al. Direct compar-ison of treatment responses, remission rates, and drugadherence in patients with rheumatoid arthritis treatedwith adalimumab, etanercept, or infliximab: results fromeight years of surveillance of clinical practice in thenationwide Danish DANBIO registry. Arthritis Rheum.2010;62:22–32.

79. Mok CC, Chan KY, Lee KL, et al on behalf of the HongKong Society of Rheumatology. Factors associated withwithdrawal of the anti-TNFα biologics in the treatmentof rheumatic diseases: data from the Hong KongBiologics Registry. Int J Rheum Dis. 2014;17(Suppl. 3):1–8.

80. Maini RN, Breedveld FC, Kalden JR, et al. Sustainedimprovement over two years in physical function, struc-tural damage, and signs and symptoms among patientswith rheumatoid arthritis treated with infliximab andmethotrexate. Arthritis Rheum. 2004;50:1051–1065.

81. Haraoui B, Cameron L, Ouellet M, et al. Anti-infliximabantibodies in patients with rheumatoid arthritis whorequire higher doses of infliximab to achieve or maintaina clinical response. J Rheumatol. 2006;33:31–36.

82. Cohen SB. Efficacy and safety of TNF inhibitors in the treat-ment of rheumatoid arthritis. Medscape 2007. Availablefrom: http://www.medscape.org/viewarticle/567983

83. Greenberg JD, Reed G, Decktor D, et al. On behalf of theCORRONA Investigators. A comparative effectivenessstudy of adalimumab, etanercept and infliximab in biolo-gically naive and switched rheumatoid arthritis patients:results from the US CORRONA registry. Ann Rheum Dis.2012;71:1134–1142.

84. Emery P. Optimizing outcomes in patients with rheuma-toid arthritis and an inadequate response to anti-TNFtreatment. Rheumatology (Oxford). 2012;51(Suppl. 5):v22–v30.

85. Jamnitski A, Bartelds GM, Nurmohamed MT, et al. Thepresence or absence of antibodies to infliximab or adali-mumab determines the outcome of switching to etaner-cept. Ann Rheum Dis. 2011;70:284–288.

• Preliminary observation that the chance of having aclinical response to a second anti-TNF agent is higherin patients who fail the first anti-TNF agent because ofanti-drug antibodies.

EXPERT OPINION ON BIOLOGICAL THERAPY 211

Dow

nloa

ded

by [

Nat

iona

l Tai

wan

Uni

vers

ity]

at 0

0:29

14

May

201

6