CliniCal pharmaCology & TherapeuTiCs 1

nature publishing group state art

In the past decade, therapy with monoclonal antibodies (mAbs) has revolutionized the therapeutic paradigms of immune diseases such as rheumatoid arthritis (RA), psoriasis, and the inflamma-tory bowel diseases (IBDs)—Crohn’s disease (CD) and ulcerative colitis (UC). The process for generating hybridomas, discovered by Köhler and Milstein in 1975, allowed for the production and isolation of mAbs as therapeutic agents.1 Early therapeutic mAbs were murine, and their use in clinical practice was limited because of immunogenicity. Subsequently, modern genetic engineering techniques led to the development of humanized and fully human mAbs. Although, in general, humanized and fully human mAbs are less immunogenic as compared with murine antibodies, they also can induce antidrug antibody (ADA) formation.

Infliximab, the first mAb used for treatment of patients with CD,2 was approved by the US Food and Drug Administration for that indication in 1998. The rapid adoption of infliximab into clinical practice, and its high level of efficacy in patients with CD unresponsive to conventional therapies, led to great interest in the development of additional mAbs to block tumor necrosis factor (TNF) activity and other proinflammatory cytokines that play key roles in the activation and perpetuation of the inflam-matory response in patients with IBD.

Even though mAbs have been used in clinical practice for more than a decade, little is known about their exposure–re-sponse relationship and the factors that may affect their dis-position. Understanding these factors is essential to further improving the therapeutic efficacy of these drugs.

This review evaluates the factors known to influence the phar-macokinetics (PK) of the currently approved mAbs for the treat-ment of IBD—infliximab, adalimumab, and certolizumab—and speculates on the future role of therapeutic drug monitoring and the role of individualized dosing for these agents.

Monoclonal antibodiesstructuremAbs are engineered immunoglobulin G (IgG) therapeutic proteins. IgG molecules are constructed by a basic unit of four polypeptide chains including two identical heavy chains (CH) and two identical light chains (CL). Each antibody comprises two domains: (i) the variable region or Fab (antigen-binding region), which is specific for the antigen target (each antibody has two Fabs), and (ii) the constant region or the Fc (Figure 1).3

Depending on their structure or isotype, mAbs can be clas-sified as murine antibodies (suffix nomenclature: -omab;

1Division of Gastroenterology, University of California San Diego, La Jolla, California, USA; 2Gastroenterology Department, Hospital Clinic of Barcelona, CIBER-EHD, IDIBAPS, University of Barcelona, Barcelona, Spain; 3Projections Research Inc., Phoenixville, Pennsylvania, USA; 4Robarts Research Institute, University of Western Ontario, London, Ontario, Canada. Correspondence: WJ Sandborn (wsandborn@ucsd.edu)

Received 4 October 2011; accepted 16 November 2011; advance online publication 22 February 2012. doi:10.1038/clpt.2011.328

Anti-TNF Monoclonal Antibodies in Inflammatory Bowel Disease: Pharmacokinetics-Based Dosing ParadigmsIngrid Ordás1,2, Diane R. Mould3, Brian G. Feagan4 and William J. Sandborn1

Crohn’s disease and ulcerative colitis are chronic inflammatory disorders resulting from immune dysregulation. patients who fail conventional medical therapy require biological treatment with monoclonal antibodies (mabs). although mabs are highly effective for induction and maintenance of clinical remission, not all patients respond, and a high proportion of patients lose response over time. one factor associated with loss of response is immunogenicity, whereby the production of antidrug antibodies accelerates mab clearance. however, other factors related to patient and disease characteristics also influence the pharmacokinetics of mabs. These factors include gender, body size, concomitant use of immunosuppressive agents, disease type, serum albumin concentration, and the degree of systemic inflammation. Because it is important to maintain clinically effective concentrations to provide optimal clinical response and drug exposure is affected by patient factors, a better understanding of the pharmacology of mabs will ultimately result in better patient care.

2 www.nature.com/cpt

state artstate art

e.g., tositumomab), chimeric (-ximab; e.g., infliximab), human-ized (-zumab; e.g., certolizumab), or “fully human” (-umab; e.g., adalimumab) (Supplementary Figure S1 online).

Infliximab is a chimeric IgG1 mAb composed of a variable murine Fab region linked by disulfide bonds to a human IgG1 κ constant region.4 Infliximab is produced by cell culture using Chinese hamster ovary cells. Adalimumab is also a recom-binant IgG1 mAb, but, unlike infliximab, it is fully human. It is composed of human-derived variable regions (Fabs) and a human IgG1 κ constant region.5 Adalimumab is produced by cell culture using Chinese hamster ovary cells. Certolizumab pegol, a pegylated humanized mAb, is composed of an IgG4 isotype Fab fragment chemically linked to a polyethylene glycol moiety, which slows clearance, thereby increasing the half-life of the drug.6 Certolizumab is produced by cell culture using Escherichia coli.

Mechanism of action and therapeutic targetImproved understanding of the pathogenesis of RA, psoriasis, and IBD at the cellular and molecular level led to the devel-opment of biological agents targeting TNF-α, a potent proin-flammatory cytokine that plays a key role in orchestrating the inflammatory process in multiple autoimmune diseases. IBD is characterized by a dysregulated mucosal immune response toward the commensal enteric flora in genetically susceptible individuals.7 This immune dysregulation results in an overpro-duction of TNF-α by monocytes, macrophages, and T cells.8 Interestingly, mAbs targeting TNF-α (infliximab, adalimumab, and certolizumab) induce the formation of regulatory macro-phages with immunosuppressive properties. This population

of macrophages inhibits proliferation of activated T cells and produces anti-inflammatory cytokines.9

TNF-α can be detected in serum in its soluble form or expressed as a cell-surface polypeptide on activated macrophages, mono-cytes, and T cells in its transmembrane form.

Overall, infliximab, adalimumab, and certolizumab have simi-lar intrinsic binding properties and affinities for both soluble and transmembrane forms of TNF-α.10 Infliximab and adalimu-mab also have similar ability to mediate complement-dependent cytotoxicity and antibody-dependent cell-mediated cytotoxicity. By contrast, certolizumab, because of the absence of the IgG1 Fc portion, exhibits neither complement-dependent cytotoxic-ity nor antibody-dependent cell-mediated cytotoxicity.10 The CH2 and CH3 domains of the IgG1 Fc portion are involved in the binding to Fc receptors of natural killer cells, which leads to the lysis of target cells. Therefore, unlike infliximab and adali-mumab, certolizumab does not induce apoptosis of activated immune cells.

Mechanisms of absorption, distribution, degradation, and eliminationAbsorption. The majority of approved mAbs are administered intravenously; however, some of these agents are designed for extravascular administration, either subcutaneous (s.c.) or intramuscular. Infliximab is administered intravenously, whereas adalimumab and certolizumab are administered by s.c. injection. In general, the route of administration of mAbs affects their pharmacokinetic behavior. Intravenous therapy allows administration of large volumes of drug, achieves imme-diate central distribution, results in less variability in drug exposure between subjects, and is usually less immunogenic. The mechanism of absorption after s.c. administration is not fully understood but is likely to occur by means of lymphatic drainage. The main drawbacks of s.c. administration are that a smaller volume must be administered (generally no more than 1 ml), in comparison with the intravenous route, and the frac-tion of dose absorbed is variable, which leads to higher phar-macokinetic variability between patients and doses. Reported bioavailability of mAbs administered subcutaneously is highly variable among individual patients, ranging from 50 to 100%.11 In addition, the s.c. route is often more immunogenic than the intravenous route because the skin is highly special-ized for processing foreign antigens. After s.c. injection, mAbs undergo slow absorption, with maximum concentrations being achieved 8 to 10 days after administration.

Distribution. After administration, mAbs distribute mainly within the central compartment (extracellular fluid), whereas penetration inside cells is limited because of their high molecu-lar weight and hydrophilicity. mAbs seem to have a volume of distribution on the order of 0.1 l/kg, approximately equal to the extracellular fluid volume.12 For example, the volume of distri-bution of infliximab at steady state ranges from 4.5 to 6 l.13

Because of the relatively large loading doses and the intra-venous route of administration, infliximab yields acute con-centration–time profiles with very high peak concentrations

SS

Fab

S S

S S

SS

Fc

VH

CH1

CH2

CH3

CH

CL

VL

Figure 1 General structure of IgG antibodies. Each antibody is composed of two identical heavy chains (CH) and two identical light chains (CL). The light chain contains two domains: one variable (VL) and one constant (CL). The heavy chain is composed of a variable domain (VH) and three constant domains (CH1, CH2, and CH3). The antigen-binding region (Fab) includes VH, VL, CL, and CH1, and the constant region (Fc) includes CH2 and CH3. -S-S-, disulfide bond.

CliniCal pharmaCology & TherapeuTiCs 3

state artstate art

and low trough levels and, hence, high peak-to-trough ratios (Figure 2).14,15 By contrast, adalimumab and certolizumab, given their slow absorption and elimination rates, exhibit more uniform concentration–time profiles at steady state.16 The phar-macokinetic properties of infliximab, adalimumab, and certoli-zumab are shown in Table 1.

Degradation and elimination. Although the exact mechanisms by which mAbs are cleared from the circulation is not well understood, the primary route of antibody clearance is via proteolytic catabolism after receptor-mediated endocytosis in the cells of the reticuloendothelial system (RES).17 Antibody salvage and recirculation is mediated by the Brambell recep-tor (FcRn), which is essential for maintaining immunoglobu-lin and albumin homeostasis.18 In adults, FcRn is primarily expressed in the vascular endothelial cells or the RES and at lower levels on monocyte cell surfaces, tissue macrophages, and dendritic cells.19 This Fc receptor plays a critical role in protecting IgG antibodies and albumin from the ongoing cat-abolic activities, thus prolonging their half-lives.20 However, this system is saturable at high IgG concentrations, resulting in an inverse relationship between concentration and half-life (the higher the concentration of the antibody, the lower its half-life).21 Thus, one would anticipate that high levels of endogenous IgG, as is seen in chronic inflammatory diseases, could potentially shorten the half-life of exogenously admin-istered mAbs.

In addition to FcRn, three other classes of Fc receptors (FcγRI, FcγII, and FcγIII) for IgG binding have been identified in humans.22 These receptors are expressed by macrophages, natural killer cells, B and T cells, and platelets. Fc gamma recep-tor (FcγR) polymorphisms have been associated with clinical response to TNF antagonists in patients with IBD.23 Whether FcγRs polymorphisms may contribute to clearance of TNF antagonists needs further investigation.

FcRn binds to IgG with pH-dependent affinity.24 Antibodies bind tightly to FcRn inside endosomes, which have an acidic environment. Inside endolysosomes, the IgG–FcRn complexes do not undergo catabolism, whereas the antibody bound to FcγR is degraded. Eventually, the antibody bound to FcRn is returned

1,000

100

10

1

0.1

Infli

xim

ab c

once

ntra

tion

(µg/

ml)

1,000

100

10

1

0.1Infli

xim

ab c

once

ntra

tion

(µg/

ml)

0 100 200

Time (days)

300 400

0 100 200

Time (days)

300 400

Model predicted infliximab concentrationObserved infliximab concentration

a

b

Figure 2 Concentration vs. time curves of infliximab in (a) ulcerative colitis (UC) and (b) Crohn’s disease (CD). Observed (dots) and simulated (lines) median concentration–time profiles of patients with UC (data from ACT trials) and CD (data from ACCENT I trial). Patients received treatment with infliximab 5 mg/kg at weeks 0, 2, and 6 and every 8 weeks thereafter. The serum infliximab concentrations are higher at early time points in the graph, corresponding with the induction phase of treatment (loading doses). After 14 weeks (100 days), infliximab serum concentrations tend to stabilize. ACCENT I, A Crohn’s Disease Clinical Trial Evaluating Infliximab in a New, Long-term Treatment Regimen; ACT, Active Ulcerative Colitis Trial. Adapted with permission from refs. 14 and 15.

table 1 Pharmacokinetic properties of infliximab, adalimumab, and certolizumab

infliximab adalimumab Certolizumab

CD uC CDa 70 uC CD71 uC

Cmax 118 µg/ml 4.7 ± 1.6 µg/ml N/A 43–49 µg/ml N/A

T1/2 7.7–9.5 days72 10–20 days N/A 14 days N/A

Tmax Within an hour 5.46 ± 2.3 days N/A 2.25–7.12 days N/A

Vd 4.5–6 liters 4.5–6 liters 4.7–6 liters N/A N/A

V1 52.4 ml/kg15 3.29 liters14

V2 19.6 ml/kg15 4.13 liters14

Cl 5.42 ml/kg/d15 (15.8 ml/h)b

0.4 liters/d14 (16.7 ml/h) 12 ml/h N/A 17 ml/h N/A

CD, Crohn’s disease; Cl, clearance; Cmax, maximum concentration; N/A, not available; T1/2, half-life; Tmax, time to reach maximum plasma concentration; UC, ulcerative colitis; V1, volume of distribution in the central compartment; V2, volume of distribution in the peripheral compartment.aFollowing a single 40-mg s.c. administration to healthy adult subjects. bAssuming a mean body weight of 70 kg.

4 www.nature.com/cpt

state artstate art

to the cell surface where, at physiologic pH, it dissociates from the receptor and is released again back into the circulation (Figure 3).

It has been shown that in mice genetically lacking expres-sion of FcRn, IgG follows hypercatabolism and thus accelerated clearance.25 These results support the importance of FcRn in regulating the catabolism of antibodies and therefore their phar-macokinetic behavior. In addition, IgG affinity for FcRn is spe-cies-specific. Human FcRn shows high affinity for human IgG, whereas the human receptor shows little affinity for IgG derived from most other species, including mice. This observation helps to explain the higher clearance that murine mAbs experience in humans.26 Thus, the structure of mAbs influences their phar-macokinetic behavior; the Fc region is responsible for the pro-longed half-life of IgG antibodies through its binding to FcRn. Fc conjugation is a common method used to extend the half-life of mAbs by decreasing their clearance through improvement of FcRn binding. As expected, the half-life of mAbs generally increases with their level of humanization. Murine antibodies display a short half-life (1–2 days); chimeric antibodies have a half-life of approximately 10 to 14 days, and humanized and fully human antibodies exhibit longer half-lives of approximately 10 to 20 days.27

PEGylation is another modification that has been used to increase the half-life of mAbs that do not have a functional Fc region. The addition of polyethylene glycol to mAbs structurally protects them from proteolytic breakdown and immunologic

recognition, thus decreasing the likelihood of neutralizing anti-body formation.

Clearance of mAbs can be either linear or nonlinear. Generally, mAbs targeting cell-surface receptors tend to exhibit nonlinear clearance that is dependent on antigen expression, whereas mAbs directed against soluble antigens (e.g., cytokines) typically exhibit dose-proportional behavior with linear clearance, which is often affected by body weight.28 However, it should be noted that the relationship between mAb clearance and body weight is generally less than linear, suggesting that mg/kg dosing may lead to patient exposure that is below the target range in patients with low body weight. There is some suggestion, however, that the clearance of mAbs targeting soluble receptors may be influ-enced by receptor expression as well.28

The contribution of receptor-mediated clearance to overall clearance depends on several factors such as mAb concentra-tion and distribution together with target receptor expression, internalization, and turnover rates. In some cases, cell-surface receptors are released into the serum, circulating as free anti-gens. mAbs can bind to these shed receptors, resulting in the formation of antibody–antigen complexes that may result in an accelerated mAb clearance.12

overview oF ibd treatMent with MabsCD and UC, the two main forms of IBD, are chronic diseases that result from immune dysregulation in genetically suscep-tible individuals.7 Because the specific cause of CD and UC is

1. mAb uptake via FcRnand FcγR interaction

Antibody

FcRn

FcγR2. Endosome formation (pH 6)

5. Cell surface (pH 7.4) Antibody is released from FcRn back into circulation

4. Antibody bound to FcRn is returned to cell surface

3. Antibody bound to FcγR is degraded and antibody bound to FcRn protected

Figure 3 Mechanism of degradation of monoclonal antibodies. FcgR, Fc gamma receptor; FcRn, Brambell receptor. Adapted with permission from ref. 54.

CliniCal pharmaCology & TherapeuTiCs 5

state artstate art

unknown, they have conventionally been treated using a broad spectrum of anti-inflammatory agents, including aminosali-cylates, corticosteroids, and standard immunosuppressive agents such as purine analogs (azathioprine and 6-mercaptopurine), methotrexate, and cyclosporine.29 However, a high proportion of patients fail to respond to these therapies and require biologi-cal treatment with TNF antagonists (mAbs). TNF antagonists have shown clear benefits in randomized controlled trials for inducing and maintaining clinical remission in both CD and UC.30–35 However, despite their therapeutic efficacy, more than one-third of patients show no response to induction therapy (primary nonresponders), and in up to 50% of responders, TNF antagonist therapy becomes ineffective over time (secondary nonresponders).36 Thus, a critical need exists to develop new approaches and strategies that further optimize the efficacy of these drugs.

In recent years, interest in the mechanistic causes of TNF antagonist treatment failure has intensified. It is now well established that patients lacking objective evidence of inflam-mation, such as presence of ulcers at the endoscopic examina-tion or increased serum C-reactive protein concentrations, do not benefit from treatment with TNF antagonists.37,38 Furthermore, symptoms usually do not correlate well with the presence of mucosal lesions or with elevated biomarkers of inflammation.39,40 Hence, when loss of response to TNF antagonists occurs, dose intensification or switching between mAbs based exclusively on symptoms will frequently lead to incorrect therapeutic decisions and should be avoided. Instead, before making any decision, patients with suspected active inflammation based on symptoms should undergo endoscopic or radiologic evaluation to ensure that they exhibit objective evidence of inflammation that could potentially ben-efit from dose intensification. A therapeutic algorithm for the treatment of patients with moderate and severe IBD is shown in Supplementary Figure S2 online. The structural charac-teristics and dosage of mAbs for the treatment of IBD are detailed in Table 2.

The development of immunogenicity due to inappropri-ate administration strategies (episodic administration rather

than scheduled administration and monotherapy instead of combined therapy with an immunosuppressive) is a common cause of treatment failure.

The optimal strategies to minimize the risk of immunogenic-ity and the potential benefits of tailoring therapy based on the determination of serum drug concentrations and the presence or absence of neutralizing ADAs are examined below.

strategies to minimize the risk of immunogenicityThe development of immunogenicity is an important determi-nant of both the efficacy and safety of TNF antagonists. As dis-cussed above, a high proportion of patients who initially respond to mAbs lose response over time, owing, in part, to development of ADAs.41–43

Two therapeutic strategies have been associated with a reduc-tion in ADA formation: (i) use of TNF antagonists in a scheduled maintenance regimen rather than episodic administration and (ii) concomitant use of immunosuppressive agents (azathioprine, mercaptopurine, or methotrexate) with a TNF antagonist.

Scheduled vs. episodic treatment. Episodic infliximab treatment strategy in patients with CD has been associated with a higher rate of antibody formation and a higher rate of infusion reac-tions as compared with scheduled maintenance therapy.44 Infusion reactions to infliximab are strongly associated with the presence of ADAs. Intravenous hydrocortisone premedica-tion reduces the formation of ADAs but does not eliminate the risk of infusion reactions.45 The best strategy to minimize this risk is to administer infliximab on a scheduled basis and to use concomitant immunosuppressive therapy.

In addition, detectable trough infliximab concentrations, which are associated with better outcomes, are higher in patients undergoing a regularly scheduled treatment strategy.44 Therefore, the optimal strategy to reduce the risk of ADAs to infliximab is to use an induction dosing regimen (5 mg/kg intravenously at weeks 0, 2, and 6) followed by a maintenance strategy every 8 weeks rather than episodic therapy. With adalimumab, episodic admin-istration has been associated with inferior clinical outcomes, but data regarding immunogenicity have not been obtained.33 With

table 2 characteristics and dosage of mabs for inflammatory bowel disease

infliximab adalimumab Certolizumab natalizumab73 ustekinumab74 golimumab75

Approved indications CD UC RA PsA AS CD UC RA PsA AS JRA Psoriasis

CD RA CD MS Psoriasis Phase II/III studies for CD

RA PsA AS Phase II/III studies for UC

Structure Chimeric IgG1 κ Human IgG1 κ Humanized pegylated Fab IgG4

Humanized IgG4 Human IgG1 κ Human IgG1 κ

Therapeutic target TNF-α TNF-α TNF-α α4 integrin IL-12/23 (p40 subunit) TNF-α

Dosage in IBD 5 mg/kg 0-2-6 and every 8 wk

160-80-40 mg/2 wk and 40 mg eow

400 mg 0-2-4 wk and every 4 wk

300 mg every 4 wk 45 mg at baseline, 4 wk after, and every 12 wk thereafter (90 mg if weight >100 kg)

50 mg monthly

Administration route i.v. s.c. s.c. i.v. s.c. s.c.

Brand name Remicade Humira Cimzia Tysabri Stelara Simponi

AS, ankylosing spondylitis; CD, Crohn’s disease; eow, every other week; Fab, antigen-binding region; IBD, inflammatory bowel disease; IgG, immunoglobulin G; IL, interleukin; i.v., intravenous; JRA, juvenile rheumatoid arthritis; mAbs, monoclonal antibodies; MS, multiple sclerosis; PsA, psoriatic arthritis; RA, rheumatoid arthritis; s.c., subcutaneous; TNF-α, tumor necrosis factor–α; UC, ulcerative colitis; wk, weeks.

6 www.nature.com/cpt

state artstate art

certolizumab, episodic treatment strategy in patients with CD has been associated with a higher rate of antibody formation as compared with scheduled maintenance therapy.32

Combined therapy vs. monotherapy. Combination therapy with TNF antagonists and immunosuppressive agents has been shown to reduce the risk of ADA formation in patients with CD using episodic or scheduled treatment strategy; the mag-nitude of reduction is amplified when the TNF antagonist is administered following a scheduled strategy.37,41,46,47 Baert et al. showed that the rate of ADA formation in patients with refractory CD treated with infliximab episodically was sig-nificantly lower in those receiving concomitant immunosup-pressive therapy as compared with those who were not taking these agents (43% vs.75%, respectively; P < 0.01).41 Vermeire et al. reproduced those results in a prospective cohort study evaluating the effectiveness of concomitant immunosuppres-sive therapy in suppressing the formation of ADAs in patients with CD treated with infliximab in an on-demand schedule. Concomitant use of azathioprine or methotrexate was associ-ated with a lower incidence of ADAs as compared with patients not taking immunosuppressive agents (46% vs. 73%, respec-tively; P < 0.001). No difference was found between azathio-prine (48%) and methotrexate (44%) in reducing this risk.46 Hanauer et al. demonstrated that concomitant use of immu-nosuppressives in patients receiving infliximab in a scheduled strategy significantly reduces the risk of immunogenicity as compared with patients receiving infliximab monotherapy (10% vs. 18%, respectively; P = 0.02).47 The SONIC trial—in which patients with CD who were naive to immunosuppres-sive and TNF antagonist therapy were randomized to receive azathioprine or infliximab or a combination of both—demon-strated that the rate of ADA formation was significantly lower in the subgroup of patients receiving combination therapy (0.9%) as compared with patients receiving infliximab mon-otherapy (14.6%).37 Feagan et al. also demonstrated that the use of concomitant methotrexate with infliximab in patients with CD significantly reduces the rate of ADA formation (4% in patients receiving combined therapy vs. 20.4% in patients treated with infliximab monotherapy—following a scheduled strategy).48

Adalimumab and certolizumab can also induce the formation of neutralizing antibodies.32,42,49 Similarly to infliximab, this risk is also decreased when these mAbs are given concomitantly with immunosuppressives and administered as a scheduled mainte-nance regimen.32,49 In patients with UC, combined immunosup-pression with azathioprine and infliximab also reduces ADA formation and increases infliximab trough concentrations.31,50,51 It is therefore clear that concomitant use of immunosuppressive agents with a TNF antagonist reduces the risk of ADA develop-ment in patients with CD and UC.

In terms of efficacy, it has recently been shown that com-bined therapy with infliximab and azathioprine is more effica-cious than either drug alone for induction of clinical remission and mucosal healing in both CD and UC.37,51 It is likely that, at least in part, this increased efficacy is due to lower rates of ADA

formation and higher infliximab drug concentrations among patients receiving combined therapy.

therapeutic monitoring: determination of trough concentra-tions and adasLoss of response to TNF antagonists, mainly due to development of neutralizing ADAs and subtherapeutic drug concentrations, is a challenging problem in the management of patients with IBD. Emerging data indicate that a strong relationship exists between serum drug concentrations (PK) and efficacy (pharmacodynam-ics, PD). Studies conducted in both RA and IBD have shown that patients with higher trough drug concentrations achieve superior outcomes.43,44,52 This observation holds out the pos-sibility that therapeutic drug monitoring may direct dose adjust-ment and clinical decision making. Infliximab concentrations ≥12 µg/ml 4 weeks after infusion or >1.4 µg/ml at dosing trough are considered to be predictive of therapeutic response.41 These cutoff values are based on a study in which patients were treated using infliximab episodically rather than on a scheduled basis and have not been prospectively validated. In a retrospective study, Afif et al. evaluated the clinical utility of measuring ADAs and trough drug concentrations in patients with loss of response to infliximab.53 In patients with antibodies against infliximab, switching to another TNF antagonist was associated with a com-plete or partial response in a very high proportion of patients (92%), whereas increasing infliximab dose had a response in only 17% of patients. Conversely, dose escalation in patients with subtherapeutic infliximab concentrations was associated with clinical response in 86% of patients, whereas the rate of clinical response in patients changing to another TNF antagonist agent was 33%. Therefore, increasing the dose of infliximab in patients who have developed ADAs is ineffective. Accordingly, measurement of ADAs and trough drug concentrations in patients with loss of response is potentially a clinically useful strategy (Supplementary Figure S2 online). Nevertheless, the added value of tailoring TNF antagonist maintenance therapy in individual patients based on trough drug concentrations and the presence or absence of ADAs deserves further evaluation by prospective studies.

Factors aFFecting the Pk and Pd oF MabsThe pharmacology of therapeutic mAbs is complex and depends not only on the structure of the antibody but also on the prop-erties of the target antigen and on patient- and disease-related factors. To date, only limited information exists regarding the factors, other than the formation of neutralizing antibodies, that influence the PK of mAbs. Identification of factors that influence disposition and elimination of mAbs is essential to understand-ing their PK–PD relationship.

PK is a branch of pharmacology dedicated to study the mecha-nisms of absorption, distribution, metabolism, and elimination of an administered drug (i.e., what the body does to the drug), whereas PD studies the relationship between drug exposure and therapeutic effect (i.e., what the drug does to the body). PK and PD are interrelated. PK–PD analyses play an important role dur-ing drug development and are critical to select the appropriate

CliniCal pharmaCology & TherapeuTiCs 7

state artstate art

dose regimens.54 Furthermore, PK–PD modeling analyses are essential to understanding the relationship between drug con-centration (PK) and therapeutic response (PD). PK–PD mod-eling is a valuable tool in drug research and development for several reasons: (i) it may help to reduce the number of unneces-sary and unproductive studies, (ii) it may help generate pivotal decision-making algorithms (e.g., dose optimization), (iii) it may help to improve the overall drug safety and efficacy, and (iv) it may lead to savings in time and money.55 PK–PD modeling has evolved rapidly over the past decade, but unfortunately no formal prospective studies evaluating the PK–PD relationship of mAbs in patients with IBD have been conducted.

Serum mAb concentrations have been shown to be highly variable between individuals and differ over time even within an individual patient. The differences in the observed concen-tration–time profiles and the exposure characteristics among mAbs can be explained by different molecular properties (such as structure, physiology of the therapeutic target, and clear-ance mechanisms), differences in the dosage or administration regimens (e.g., route of administration and administration fre-quency), and patient and disease characteristics. These factors are discussed below.

Population PK studies seek to identify the factors that influ-ence changes in the relationship between the administered dose and the achieved serum concentrations. Hence, if therapeutic concentrations are not reached, dosage can be appropriately modified according to these factors. Unfortunately, several uncertainties regarding the pharmacokinetic profile of mAbs persist. Specifically, the covariates that may influence drug clear-ance are not well defined.

Considering trough serum drug concentrations as a surrogate for pharmacokinetic analysis may be misleading and related to errors. Estimates from nontrough PK observations are more robust, providing more information on the disposition of the drug, and are thus preferred for PK–PD modeling studies. Population PK models, using random variables with mean and variance parameters instead of an individual data analysis, help to identify and quantify the sources of variability through the identification of covariates on each pharmacokinetic parameter. In addition, population PK analyses supported by large data sets with sparse sampling yield good-quality pharmacokinetic parameter estimates and can be better extrapolated to the tar-get patient population as compared with the values obtained from studies involving single-dose administrations and a small number of subjects because the results obtained from population PK analyses reflect information from a wide range of patients undergoing treatment at multiple clinical centers. Therefore, PK modeling studies using intensive blood sampling instead of measuring only trough drug concentrations, together with evaluation of several patient and disease covariates, are more accurate and yield more information on the sources of between-patient variability in mAb exposure.

Although therapeutic mAbs have been commercially available for two decades, little is known about their PK–PD relationship. Conventional wisdom implicates neutralizing ADAs as a pri-mary cause of therapeutic failure. However, although ADAs can

profoundly affect drug clearance, resulting in low or nonmeas-urable trough drug concentrations and loss of response, other factors that affect the PK of TNF antagonists exist, including con-comitant use of immunosuppressives, serum albumin concen-tration, body weight, the degree of systemic inflammation (e.g., serum albumin concentration and TNF burden), and disease type (e.g., CD vs. UC). Collectively, these factors probably account for the large interindividual differences in PK and clinical efficacy observed after standard dosing of mAbs (Table 3).

determinants of the Pk–Pd of mabs: challenges in interpret-ing the literatureIt should be noted that, to date, the majority of publications evaluating the relationship between the PK and PD of mAbs have been compromised by the following problems: (i) retro-spective study designs that are not optimally designed to identify relevant PK–PD relationships; (ii) failure to accurately sample serum at the time of treatment failure/success; (iii) failure to per-form pharmacokinetic sampling at informative times, thus limit-ing analytical power (peak and concentrations measured during the beta decline phase are more informative of drug clearance than trough samples); (iv) use of ADA assays that cannot detect ADAs in the presence of circulating drug; (v) use of inappropri-ate statistical methods (“as observed” analyses do not adequately account for patients who withdraw from treatment prematurely and who are therefore more likely to have measurable ADAs and low serum drug concentrations; intent-to-treat evaluations suf-fer from other issues);56 (vi) failure to account analytically for the effects of confounders; (vii) inclusion of patients without evidence of active inflammation, thus reducing statistical power; and (viii) failure to use objective PD end points (the majority of studies have used symptoms instead of objective findings of inflammation).

Factors that may influence the PK and hence the PD of mAbs in patients with IBD are reviewed below.

table 3 Factors affecting the pharmacokinetics of monoclonal antibodies

impact on pharmacokinetics

Presence of ADAs Decreases serum (mAbs) Threefold-increased clearance Worse clinical outcomes

Concomitant use of IS Reduces ADA formation Increases serum (mAbs) Decreases mAbs clearance Better clinical outcomes

High baseline (TNF-α) May decrease (mAbs) by increasing clearance

Low albumin Increases clearance Worse clinical outcomes

High baseline CRP Increases clearance

Body size High body mass index may increase clearance

Gender Males have higher clearance

ADA, antidrug antibody; CRP, C-reactive protein; IS, immunosuppressive agent; mAb, monoclonal antibody; TNF-α, tumor necrosis factor-α. Terms in parentheses refer to serum concentration.

8 www.nature.com/cpt

state artstate art

Role of ADAs. Immunogenicity is a major issue related to thera-peutic efficacy of mAbs. Development of ADAs can affect the safety profile of these drugs because of hypersensitivity reactions. mAbs are exogenous proteins and can therefore induce an immune response leading to the production of endogenous ADAs, which in turn leads to a reduced therapeu-tic efficacy of the drug. Depending on the structure of mAbs, ADAs can be classified as human antimouse antibodies, human antichimeric antibodies, or human antihuman antibodies. Theoretically, immunogenicity decreases with the degree of humanization. Although infliximab, because of its chimeric structure, is theoretically more immunogenic than adalimu-mab and certolizumab, data from clinical trials evaluating the efficacy of these mAbs in patients with IBD confirm that both adalimumab and certolizumab can also induce immunogenic-ity and that the degree of immunogenicity seems to be rela-tively similar to that seen with infliximab.32,42,49,57

Development of ADAs results in the formation of immune complexes that accelerate drug clearance by the RES and/or impaired binding to target.

If the drug–antibody complex is inactive (“neutralizing antibody”) and if the antibody binding capacity is similar to the total concen-tration of the therapeutic protein, decreased efficacy may ensue as a result of a decline in free concentrations of the active agent. Alternatively, if the drug–antibody complex is active, enhanced bio-activity may result.55 In some cases, the drug–antibody complex will be cleared (“clearing antibody”), which provides an alternative clearance pathway for the therapeutic protein, decreasing total and free concentrations and leading to decreased bioactivity.

Therefore, immunogenicity usually impacts clinical response to therapy in a negative manner by affecting bioavailability, PK, and PD.58 However, this topic has been controversial because of methodological challenges. The majority of published data evaluating the influence of ADAs on the pharmacokinetic properties of mAbs are based on solid-phase enzyme-linked immunosorbent assays in which the presence of circulating drug renders the test insensitive in detecting ADAs. In addi-tion, solid-phase enzyme-linked immunosorbent assay are associated with false-positive results due to nonspecific binding to immunoglobulins other than infliximab. However, highly sensitive liquid-phase mobility-shift assays and liquid-phase radioimmunoassays that measure ADAs in the presence of cir-culating drug are emerging, which should provide more accu-rate evaluation of the rate and intensity of sensitization early in the course of treatment with mAbs. This advance highlights the potential of therapeutic drug monitoring to direct interven-tions, such as dose intensification or immunosuppression, that may prevent primary and secondary loss of response.

Several studies have consistently linked the presence of ADAs to inferior outcomes.37,41,42 Baert et al. showed that the presence of a high titer of ADAs in patients with CD was associated with a reduced duration of response in comparison with nonsensi-tized patients (35 days vs. 71 days, respectively; P < 0.001).41 Similarly, the formation of ADAs in patients with CD has been associated with lower rates of prednisone-free clinical remis-sion (57.1% (prednisone-free clinical remission in patients with

positive ADAs) vs. 70.6% (prednisone-free clinical remission in patients with negative ADAs), respectively).37

Role of concomitant immunosuppressive therapy. Although the mechanism by which immunosuppressives (azathioprine, mer-captopurine, and methotrexate) increase the concentration of serum mAbs is not well established, they are likely to exert this function by reducing the formation of ADAs and/or downregu-lating RES-mediated drug clearance. In some cases, immuno-suppressives downregulate receptors for mAbs, which also slows the clearance of mAbs. Because of the methodological problems previously described, the role of concomitant immunosup-pressive therapy as a determinant of the PK of mAbs is poorly understood. Post hoc analysis of four randomized controlled trials showed that concomitant use of immunosuppressives with infliximab was associated with higher serum infliximab concentrations.50 In the SONIC trial, patients with active CD who received combination therapy (infliximab plus azathio-prine) had higher trough infliximab concentrations than those who received infliximab monotherapy (3.5 μg/ml vs. 1.6 μg/ml, respectively; P < 0.001). These findings correlated with bet-ter outcomes in terms of higher corticosteroid-free remission rates in the combination therapy arm.37 Although it is apparent that coadministration of azathioprine decreased drug clearance in the SONIC trial, the mechanisms responsible are unclear. One likely mechanism is reduction of ADA formation (0.9% in patients receiving combination therapy vs. 14.6% in patients receiving infliximab monotherapy).

As mentioned above, coadministration of immunosuppres-sives with TNF antagonists increases serum drug concentrations and decreases the formation of ADAs.41,46,47,50 In patients with UC, however, factors other than immunosuppressive-mediated clearance may have, quantitatively, the most critical effect in determining the PK of infliximab, at least during the acute phase of therapy.

Role of the RES and disease severity. Because of their high molecular weight, mAbs do not undergo renal elimination or metabolism by hepatic enzymes; rather, as previously described, proteolytic catabolism within the cells of the RES is the primary route of elimination.12 Disease severity may influence elimination of mAbs through RES-mediated mechanisms. In regard to this observation, it has been shown that patients with elevated C-reactive protein and serum albumin concentrations below the normal range have accelerated drug clearance.14,15,59 The presence of systemic inflammation may therefore increase mAbs catabolism in the RES. Unfortunately, there is no practi-cal way to measure this phenomenon. However, generating data for each mAb and disease using a range of covariates related to the inflammatory burden of the disease could be an indirect way to determine the degree of systemic inflammation required before the PK of mAbs is substantially affected.

This assumption holds out the possibility that patients with more severe inflammation may require higher than average drug exposure for optimal results. This hypothesis may account for the suboptimal infliximab concentrations that have been observed

CliniCal pharmaCology & TherapeuTiCs 9

state artstate art

in patients with severe UC undergoing infliximab induction therapy.43

As an additional route of mAb clearance, the intestinal clear-ance of IgG was investigated in patients with IBD (inactive and active CD and UC) as well as in a control group. The authors reported increases in intestinal clearance of monomeric IgG that were closely related to the severity of the intestinal lesions.60

Role of the “antigen sink”. The antigen-dependent clearance path-way is often referred as an “antigen sink.”

It has been shown that receptor density (i.e., antigen targeted by mAbs) clearly influences the PK of mAbs. Therefore, differences in PK do exist with different indications, and, indeed, changes in PK with patient response have been reported. In the case of gemtuzu-mab, an antibody targeting CD33-positive blast cells with chemo-therapeutic properties in patients with acute myeloid leukemia, serum concentrations and half-life increased after a second dose as compared with the first dose of the drug. This observation was attributed to a decreased clearance by CD33-positive blast cells due to a reduced tumor burden after the first dose.61

One potential explanation for lack of response to TNF antag-onists is incomplete suppression of TNF-α activity because of insufficient serum drug concentrations to block the excess of TNF. A high inflammatory burden at baseline is associated with higher concentrations of TNF-α in both tissue and serum. Patients with a higher degree of systemic inflammation may therefore require, in a stoichiometric fashion, greater amounts of drug to neutralize this excess of TNF-α. In turn, this could result in lower mAb serum concentrations and less functional available drug. In this paradigm, the higher the baseline TNF-α concentration, the higher the dose of drug required to achieve a pharmacodynamic response. In addition, baseline TNF-α serum concentrations may predict the need of dose escalation in cases of loss of response.62 Ainsworth et al. evaluated 33 patients with CD treated with infliximab classified by response status (primary nonresponse, secondary loss of response, and sustained response) and found that patients who were primary nonresponders to infliximab had higher affinity to bind TNF-α in serum than patients with secondary loss of response and had no antibod-ies against infliximab (Supplementary Figure S3 online). The authors concluded that measurement of serum TNF-α binding capacity in conjunction with ADAs may provide new insights into the causes of treatment failure (sensitization vs. non-TNF inflammatory pathway vs. inadequate drug concentration in the absence of sensitization).63 Interestingly, not only can the meas-urement of serum TNF-α concentration be used as a surrogate marker of the PK of mAbs but its measurement in colonic tissue seems to also be useful for this purpose. Olsen et al. demonstrated that the likelihood of inducing clinical or endoscopic remission after induction therapy with infliximab in patients with UC was inversely associated with pretreatment concentration of TNF-α in colorectal mucosa.64 The clinical implication of this observa-tion is that pretreatment values of colorectal TNF-α may be used, together with other factors, as a surrogate marker to individualize infliximab dosing regimen. Patients with higher pretreatment TNF-α colonic concentration may require higher doses.

Role of disease type: CD vs. UC. Potential pharmacokinetic differ-ences between CD and UC exist that may or may not result from differences between the previously discussed factors. Infliximab clearance seems to be similar for CD, RA, and pso-riasis. In distinction, potentially important pharmacokinetic differences exist between CD and UC.43,44 Seow et al. dem-onstrated that patients with moderate UC had higher rates of clinical response (70% vs. 41%; P = 0.004), clinical remission (41% vs. 17%; P = 0.015), and endoscopic remission (26% vs. 4%; P = 0.046) than those with severe disease after infliximab induction treatment. Undetectable trough infliximab concen-trations were associated with less favorable outcomes.43 In the study by Seow et al., the proportion of patients with non-measurable infliximab trough concentrations was higher than that in a previous study performed by the same investigators in patients with CD.44 One potential explanation of these find-ings is that patients with UC have a more rapid clearance of infliximab than patients with CD. It is noteworthy that two large-scale randomized con trolled trials have shown relatively low rates of remission with s.c. adalimumab induction therapy in UC35,65 at doses that result in relatively high rates of remis-sion in CD. In distinction, both intravenously administered infliximab and subcutaneously administered adalimumab are effective in CD.30,49 These findings raise the possibility that fun-damental pharmacokinetic differences with important clinical consequences may exist between the two diseases. One hypoth-esis that may account for these differences is that the overall inflammatory burden in patients with severe and extensive UC is higher than that in patients with CD owing to a greater inflamed intestinal surface. This higher inflammatory burden may result in higher production of TNF-α. Therefore, higher doses of mAbs may be required to neutralize the excessive pro-duction of the target antigen. Alternatively, patients with UC may have a greater area of the mucosal surface affected than is observed in patients with CD, resulting in greater loss of drug in the intestinal lumen. Although the concept of a “drug-losing enteropathy” is entirely hypothetical, future pharmacokinetic studies should explore this hypothesis.

Role of patient factors. It is worth noting that the influence of weight on the clearance, and therefore area under the curve, of mAbs is not linear. Hence, dosing based on weight does not always produce drug exposure that is efficacious. Monitoring of serum drug concentrations is consequently more impor-tant in patients with both low weight and high inflammatory burden than in patients with higher weight and/or less inflam-mation. Gender has also been shown to independently influ-ence the disposition of mAbs, with clearance being higher in men.14,66 However, because weight and gender are generally somewhat correlated (men weighing more than women), this finding may be related to weight.

The clearance of mAbs is not affected by either renal or hepatic dysfunction. However, an interesting association between base-line serum albumin concentrations and serum infliximab con-centrations in both UC and CD has recently been reported.15,67 Patients with a baseline serum albumin concentration below the

10 www.nature.com/cpt

state artstate art

normal range (a common finding associated with severe inflam-mation) have lower remission rates after treatment with inflixi-mab. This observation suggests that an inverse relationship exists between serum albumin concentration and infliximab clearance

(the lower the albumin concentration, the higher the infliximab clearance) (Figure 4).

Recently, a study of patients with RA treated with infliximab has shown that a high body mass index negatively influences clinical response.68 Research into the role of mesenteric fat in chronic inflammatory diseases has intersected with investigation into the importance of adipose tissue as a metabolically active source of proinflammatory cytokines (e.g., TNF) in patients with insulin resistance. It would be expected, therefore, that obese patients with CD would inherently have higher TNF produc-tion than patients with normal weight, suggesting also that the mg/ kg dose paradigm might be inappropriate for obese patients; these patients may require higher drug doses than those cur-rently recommended. This observation requires confirmation in additional studies.

Accordingly, measurement of body mass index (and poten-tially quantitative assessment of mesenteric fat) should be incorporated into future pharmacokinetic studies, especially in patients with CD, in whom adipose tissue is involved in the inflammatory process.69

In summary, preliminary evidence suggests that multiple fac-tors influence the PK of mAbs. Understanding the determinants of the PK of mAbs has great potential to improve and optimize the therapeutic management of patients with IBD.

conclusions and Future directionsThe use of anti-TNF mAbs in the treatment of IBD has led to improved disease outcomes. However, further optimization is needed because a high proportion of patients fail to respond to these therapies. Monitoring serum drug concentrations and ADAs (immunogenicity) may lead to more appropriate thera-peutic management of patients with loss of response.

The PK of mAbs seems to be strongly influenced by several factors related to patient and disease characteristics. Evaluation of the covariates that influence the disposition of mAbs may help in identifying patients who are more likely to benefit from receiving higher doses as a result of accelerated drug clearance. Unfortunately, studies integrating these variables into a single PK model in targeted populations have not been performed yet.

The number of approved mAbs for the treatment of IBD is expected to increase. Therefore, a better understanding of the factors that impact the PK and PD of mAbs is crucial to ensure more efficient dosing regimens, which in turn may enhance the therapeutic success of these therapies. Combined clini-cal, imaging, and PK studies should lead to further advances in customizing drug dosage and monitoring therapeutic response. Finally, individualized and tailored dosing approaches guided by PK algorithms may be safer, more effective, and even cost-effective.

suPPleMentarY Material is linked to the online version of the paper at http://www.nature.com/cpt

conFlict oF interestThe authors declare no financial or other conflict of interest in relation to the content of this article. I.O. does not have any stocks, equity, a contract of employment, or a named position on a company board in companies

−1.0

−0.5

0.0

0.5

1.0

CL

(ml/k

g/d)

Base model

Base model

2.5 3.0 3.5 4.0

SAC (g/dl)

4.5 5.0

20

15

10

5

CL

(ml/k

g/d)

2 3

SAC (g/dl)

4 5

0

20

40

% R

espo

nder

s 60

80

Albumin level(g/dl)

<3.5 ≥3.5Placebo

<3.5 ≥3.5IFX-Treated

All albumin concentrationsPlacebo IFX-Treated

P = 0.2236

P = 0.0021 P = 0.0001

a

b

c

Figure 4 Albumin as a predictive factor of infliximab clearance. Serum albumin concentration (SAC) is inversely related to infliximab clearance (CL) in both (a) ulcerative colitis (UC) and (b) Crohn’s disease. (c) Relationship between serum album concentrations and clinical response rates in patients with UC treated with infliximab (IFX) and placebo. Patients with serum albumin concentration below the normal range achieved lower response rates. Adapted with permission from refs. 15 and 67.

CliniCal pharmaCology & TherapeuTiCs 11

state artstate art

related to IBD; does not hold any relevant patents that are licensed; and does not have any research support, lecture fees, or consultancies related to IBD. D.R.M. has been a paid consultant for Centocor but does not own any stock in this company. B.G.F. has received research support consulting and lecture fees from Janssen (previously Centocor), Merck (previously Schering Plough), Abbott Laboratories, and UCB Pharma. W.J.S. has received research support and consulting fees from Janssen (previously Centocor), Merck (previously Schering Plough), Abbott Laboratories, and UCB Pharma and has received lecture fees from Janssen and Abbott Laboratories

© 2012 American Society for Clinical Pharmacology and Therapeutics

1. Köhler, G. & Milstein, C. Continuous cultures of fused cells secreting antibody of predefined specificity. Nature 256, 495–497 (1975).

2. Present, D.H. et al. Infliximab for the treatment of fistulas in patients with Crohn’s disease. N. Engl. J. Med. 340, 1398–1405 (1999).

3. Edelman, G.M. Antibody structure and molecular immunology. Science 180, 830–840 (1973).

4. Knight, D.M. et al. Construction and initial characterization of a mouse-human chimeric anti-TNF antibody. Mol. Immunol. 30, 1443–1453 (1993).

5. Salfeld, J. et al. Human antibodies that bind human TNFα. US patent 6090382 (2000).

6. Athwal D. et al. Biological products. US patent 7012135 (2006).7. Xavier, R.J. & Podolsky, D.K. Unravelling the pathogenesis of inflammatory

bowel disease. Nature 448, 427–434 (2007).8. Reinecker, H.C. et al. Enhanced secretion of tumour necrosis factor-alpha, IL-6,

and IL-1 beta by isolated lamina propria mononuclear cells from patients with ulcerative colitis and Crohn’s disease. Clin. Exp. Immunol. 94, 174–181 (1993).

9. Vos, A.C., Wildenberg, M.E., Duijvestein, M., Verhaar, A.P., van den Brink, G.R. & Hommes, D.W. Anti-tumor necrosis factor-a antibodies induce regulatory macrophages in an Fc region-dependent manner. Gastroenterol. 140, 221–230 (2011).

10. Nesbitt, A. et al. Mechanism of action of certolizumab pegol (CDP870): in vitro comparison with other anti-tumor necrosis factor alpha agents. Inflamm. Bowel Dis. 13, 1323–1332 (2007).

11. Lobo, E.D., Hansen, R.J. & Balthasar, J.P. Antibody pharmacokinetics and pharmacodynamics. J. Pharm. Sci. 93, 2645–2668 (2004).

12. Mould, D.R. & Green, B. Pharmacokinetics and pharmacodynamics of monoclonal antibodies: concepts and lessons for drug development. BioDrugs 24, 23–39 (2010).

13. Klotz, U., Teml, A. & Schwab, M. Clinical pharmacokinetics and use of infliximab. Clin. Pharmacokinet. 46, 645–660 (2007).

14. Fasanmade, A.A. et al. Population pharmacokinetic analysis of infliximab in patients with ulcerative colitis. Eur. J. Clin. Pharmacol. 65, 1211–1228 (2009).

15. Fasanmade, A.A., Adedokun, O.J., Blank, M., Zhou, H. & Davis, H.M. Pharmacokinetic properties of infliximab in children and adults with Crohn’s disease: a retrospective analysis of data from 2 phase III clinical trials. Clin. Ther. 33, 946–964 (2011).

16. Nestorov, I. Clinical pharmacokinetics of TNF antagonists: how do they differ? Semin. Arthritis Rheum. 34, 12–18 (2005).

17. Wang, W., Wang, E.Q. & Balthasar, J.P. Monoclonal antibody pharmacokinetics and pharmacodynamics. Clin. Pharmacol. Ther. 84, 548–558 (2008).

18. Brambell, F.W., Hemmings, W.A. & Morris, I.G. A theoretical model of gamma-globulin catabolism. Nature 203, 1352–1354 (1964).

19. Zhu, X. et al. MHC class I-related neonatal Fc receptor for IgG is functionally expressed in monocytes, intestinal macrophages, and dendritic cells. J. Immunol. 166, 3266–3276 (2001).

20. Telleman, P. & Junghans, R.P. The role of the Brambell receptor (FcRB) in liver: protection of endocytosed immunoglobulin G (IgG) from catabolism in hepatocytes rather than transport of IgG to bile. Immunol. 100, 245–251 (2000).

21. Morell, A., Terry, W.D. & Waldmann, T.A. Metabolic properties of IgG subclasses in man. J. Clin. Invest. 49, 673–680 (1970).

22. Cohen-Solal, J.F., Cassard, L., Fridman, W.H. & Sautès-Fridman, C. Fc gamma receptors. Immunol. Lett. 92, 199–205 (2004).

23. Louis, E. et al. Association between polymorphism in IgG Fc receptor IIIa coding gene and biological response to infliximab in Crohn’s disease. Aliment. Pharmacol. Ther. 19, 511–519 (2004).

24. Junghans, R.P. Finally! The Brambell receptor (FcRB). Mediator of transmission of immunity and protection from catabolism for IgG. Immunol. Res. 16, 29–57 (1997).

25. Israel, E.J., Wilsker, D.F., Hayes, K.C., Schoenfeld, D. & Simister, N.E. Increased clearance of IgG in mice that lack beta 2-microglobulin: possible protective role of FcRn. Immunology 89, 573–578 (1996).

26. Ober, R.J., Radu, C.G., Ghetie, V. & Ward, E.S. Differences in promiscuity for antibody-FcRn interactions across species: implications for therapeutic antibodies. Int. Immunol. 13, 1551–1559 (2001).

27. Ternant, D. & Paintaud, G. Pharmacokinetics and concentration-effect relationships of therapeutic monoclonal antibodies and fusion proteins. Expert Opin. Biol. Ther. 5 (suppl. 1), S37–S47 (2005).

28. Tabrizi, M.A., Tseng, C.M. & Roskos, L.K. Elimination mechanisms of therapeutic monoclonal antibodies. Drug Discov. Today 11, 81–88 (2006).

29. Travis, S.P. et al.; European Crohn’s and Colitis Organisation. European evidence based consensus on the diagnosis and management of Crohn’s disease: current management. Gut 55 (suppl. 1), i16–i35 (2006).

30. Hanauer, S.B. et al.; ACCENT I Study Group. Maintenance infliximab for Crohn’s disease: the ACCENT I randomised trial. Lancet 359, 1541–1549 (2002).

31. Rutgeerts, P. et al. Infliximab for induction and maintenance therapy for ulcerative colitis. N. Engl. J. Med. 353, 2462–2476 (2005).

32. Schreiber, S. et al.; PRECISE 2 Study Investigators. Maintenance therapy with certolizumab pegol for Crohn’s disease. N. Engl. J. Med. 357, 239–250 (2007).

33. Colombel, J.F. et al. Adalimumab for maintenance of clinical response and remission in patients with Crohn’s disease: the CHARM trial. Gastroenterol. 132, 52–65 (2007).

34. Sandborn, W.J. et al.; PRECISE 1 Study Investigators. Certolizumab pegol for the treatment of Crohn’s disease. N. Engl. J. Med. 357, 228–238 (2007).

35. Reinisch, W. et al. Adalimumab for induction of clinical remission in moderately to severely active ulcerative colitis: results of a randomised controlled trial. Gut 60, 780–787 (2011).

36. Peyrin-Biroulet, L., Deltenre, P., de Suray, N., Branche, J., Sandborn, W.J. & Colombel, J.F. Efficacy and safety of tumor necrosis factor antagonists in Crohn’s disease: meta-analysis of placebo-controlled trials. Clin. Gastroenterol. Hepatol. 6, 644–653 (2008).

37. Colombel, J.F. et al.; SONIC Study Group. Infliximab, azathioprine, or combination therapy for Crohn’s disease. N. Engl. J. Med. 362, 1383–1395 (2010).

38. Bruining, D.H. & Sandborn, W.J. Do not assume symptoms indicate failure of anti-tumor necrosis factor therapy in Crohn’s disease. Clin. Gastroenterol. Hepatol. 9, 395–399 (2011).

39. Modigliani, R. et al. Clinical, biological, and endoscopic picture of attacks of Crohn’s disease. Evolution on prednisolone. Groupe d’Etude Thérapeutique des Affections Inflammatoires Digestives. Gastroenterol. 98, 811–818 (1990).

40. Jones, J. et al. Relationships between disease activity and serum and fecal biomarkers in patients with Crohn’s disease. Clin. Gastroenterol. Hepatol. 6, 1218–1224 (2008).

41. Baert, F. et al. Influence of immunogenicity on the long-term efficacy of infliximab in Crohn’s disease. N. Engl. J. Med. 348, 601–608 (2003).

42. Karmiris, K. et al. Influence of trough serum levels and immunogenicity on long-term outcome of adalimumab therapy in Crohn’s disease. Gastroenterol. 137, 1628–1640 (2009).

43. Seow, C.H., Newman, A., Irwin, S.P., Steinhart, A.H., Silverberg, M.S. & Greenberg, G.R. Trough serum infliximab: a predictive factor of clinical outcome for infliximab treatment in acute ulcerative colitis. Gut 59, 49–54 (2010).

44. Maser, E.A., Villela, R., Silverberg, M.S. & Greenberg, G.R. Association of trough serum infliximab to clinical outcome after scheduled maintenance treatment for Crohn’s disease. Clin. Gastroenterol. Hepatol. 4, 1248–1254 (2006).

45. Farrell, R.J., Alsahli, M., Jeen, Y.T., Falchuk, K.R., Peppercorn, M.A. & Michetti, P. Intravenous hydrocortisone premedication reduces antibodies to infliximab in Crohn’s disease: a randomized controlled trial. Gastroenterol. 124, 917–924 (2003).

46. Vermeire, S., Noman, M., Van Assche, G., Baert, F., D’Haens, G. & Rutgeerts, P. Effectiveness of concomitant immunosuppressive therapy in suppressing the formation of antibodies to infliximab in Crohn’s disease. Gut 56, 1226–1231 (2007).

47. Hanauer, S.B. et al. Incidence and importance of antibody responses to infliximab after maintenance or episodic treatment in Crohn’s disease. Clin. Gastroenterol. Hepatol. 2, 542–553 (2004).

48. Feagan, B.G. et al. Methotrexate for the prevention of antibodies to infliximab in patients with Crohn´s disease (abstract S1051). Presented at Digestive Disease Week, New Orleans, LA, 2 May 2010.

49. Sandborn, W.J. et al. Adalimumab for maintenance treatment of Crohn’s disease: results of the CLASSIC II trial. Gut 56, 1232–1239 (2007).

50. Lichtenstein, G.R. et al. Clinical trial: benefits and risks of immunomodulators and maintenance infliximab for IBD-subgroup analyses across four randomized trials. Aliment. Pharmacol. Ther. 30, 210–226 (2009).

12 www.nature.com/cpt

state artstate art

51. Panaccione, R. et al. Infliximab, azathioprine, or infliximab + azathioprine for treatment of moderate to severe ulcerative colitis: The UC SUCCESS trial. J Crohn’s and Colitis 5, S3–S12. (2011).

52. Radstake, T.R. et al. Formation of antibodies against infliximab and adalimumab strongly correlates with functional drug levels and clinical responses in rheumatoid arthritis. Ann. Rheum. Dis. 68, 1739–1745 (2009).

53. Afif, W. et al. Clinical utility of measuring infliximab and human anti-chimeric antibody concentrations in patients with inflammatory bowel disease. Am. J. Gastroenterol. 105, 1133–1139 (2010).

54. Mould, D.R. & Sweeney, K.R. The pharmacokinetics and pharmacodynamics of monoclonal antibodies–mechanistic modeling applied to drug development. Curr. Opin. Drug Discov. Devel. 10, 84–96 (2007).

55. Galluppi, G.R. et al. Integration of pharmacokinetic and pharmacodynamic studies in the discovery, development, and review of protein therapeutic agents: a conference report. Clin. Pharmacol. Ther. 69, 387–399 (2001).

56. Sheiner, L.B. Is intent-to-treat analysis always (ever) enough? Br. J. Clin. Pharmacol. 54, 203–211 (2002).

57. Schreiber, S. et al.; CDP870 Crohn’s Disease Study Group. A randomized, placebo-controlled trial of certolizumab pegol (CDP870) for treatment of Crohn’s disease. Gastroenterol. 129, 807–818 (2005).

58. Bendtzen, K., Ainsworth, M., Steenholdt, C., Thomsen, O.Ø. & Brynskov, J. Individual medicine in inflammatory bowel disease: monitoring bioavailability, pharmacokinetics and immunogenicity of anti-tumour necrosis factor-alpha antibodies. Scand. J. Gastroenterol. 44, 774–781 (2009).

59. Wolbink, G.J. et al. Relationship between serum trough infliximab levels, pretreatment C reactive protein levels, and clinical response to infliximab treatment in patients with rheumatoid arthritis. Ann. Rheum. Dis. 64, 704–707 (2005).

60. Kapel, N., Meillet, D., Favennec, L., Magne, D., Raichvarg, D. & Gobert, J.G. Evaluation of intestinal clearance and faecal excretion of alpha 1-antiproteinase and immunoglobulins during Crohn’s disease and ulcerative colitis. Eur. J. Clin. Chem. Clin. Biochem. 30, 197–202 (1992).

61. Dowell, J.A., Korth-Bradley, J., Liu, H., King, S.P. & Berger, M.S. Pharmacokinetics of gemtuzumab ozogamicin, an antibody-targeted chemotherapy agent for the treatment of patients with acute myeloid leukemia in first relapse. J. Clin. Pharmacol. 41, 1206–1214 (2001).

62. Takeuchi, T. et al. Baseline tumour necrosis factor alpha levels predict the necessity for dose escalation of infliximab therapy in patients with rheumatoid arthritis. Ann. Rheum. Dis. 70, 1208–1215 (2011).

63. Ainsworth, M.A., Bendtzen, K. & Brynskov, J. Tumor necrosis factor-alpha binding capacity and anti-infliximab antibodies measured by fluid-phase radioimmunoassays as predictors of clinical efficacy of infliximab in Crohn’s disease. Am. J. Gastroenterol. 103, 944–948 (2008).

64. Olsen, T., Goll, R., Cui, G., Christiansen, I. & Florholmen, J. TNF-alpha gene expression in colorectal mucosa as a predictor of remission after induction therapy with infliximab in ulcerative colitis. Cytokine 46, 222–227 (2009).

65. Sandborn, W.J. et al. Adalimumab induces and maintains clinical remission in patients with moderate-to-severe ulcerative colitis. Gastroenterol. 142, 257–265 (2012).

66. Ternant, D. et al. Infliximab pharmacokinetics in inflammatory bowel disease patients. Ther. Drug Monit. 30, 523–529 (2008).

67. Fasanmade, A.A., Adedokun, O.J., Olson, A., Strauss, R. & Davis, H.M. Serum albumin concentration: a predictive factor of infliximab pharmacokinetics and clinical response in patients with ulcerative colitis. Int. J. Clin. Pharmacol. Ther. 48, 297–308 (2010).

68. Klaasen, R., Wijbrandts, C.A., Gerlag, D.M. & Tak, P.P. Body mass index and clinical response to infliximab in rheumatoid arthritis. Arthritis Rheum. 63, 359–364 (2011).

69. Peyrin-Biroulet, L. et al. Mesenteric fat in Crohn’s disease: a pathogenetic hallmark or an innocent bystander? Gut 56, 577–583 (2007).

70. Adalimumab (Humira). FDA prescribing information. Revised March 2011.71. Certolizumab (Cimzia). FDA prescribing information. Revised November 2009.72. Infliximab (Remicade). FDA prescribing information. Revised February 2011.73. Ghosh, S. et al.; Natalizumab Pan-European Study Group. Natalizumab for

active Crohn’s disease. N. Engl. J. Med. 348, 24–32 (2003).74. Sandborn, W.J. et al.; Ustekinumab Crohn’s Disease Study Group. A

randomized trial of Ustekinumab, a human interleukin-12/23 monoclonal antibody, in patients with moderate-to-severe Crohn’s disease. Gastroenterol. 135, 1130–1141 (2008).

75. ClinicalTrials.gov. A Study of the Safety and Effectiveness of CNTO 148 (Golimumab) in Patients With Moderately to Severely Active Ulcerative Colitis <http://clinicaltrials.gov/ct2/show/NCT00488631>.