Secondary Logo

Journal Logo

RHEUMATOID ARTHRITIS: Edited by Ronald F. van Vollenhoven

Immunogenicity of biological therapeutics

from assay to patient

Krieckaert, Charlottea; Rispens, Theob; Wolbink, Gertjana,b

Author Information
Current Opinion in Rheumatology: May 2012 - Volume 24 - Issue 3 - p 306-311
doi: 10.1097/BOR.0b013e3283521c4e
  • Free

Abstract

INTRODUCTION

Biological medications include a wide range of medicinal products created by biological instead of chemical processes. Biologicals can consist of proteins, nucleic acids or complex combinations of substances, or may be living entities such as cells and tissues. They are isolated from natural sources or are produced by biotechnology methods [1].

Biological therapeutics can be recognised by the human immune system as ‘non self’ and induce an immune response, also known as immunogenicity. Therapeutic endogenous proteins, such as erythropoietin and growth factors, possess an amino acid sequence identical to the human equivalent and can be immunogenic due to glycosylation or conformational changes, whereby new epitopes are exposed. Therapeutic antibodies carry unique complementarity determining regions, containing stretches of sequences that frequently will be recognized as foreign and induce formation of antidrug antibodies (ADAs).

Controversy exists on the clinical impact of immunogenicity of biological therapeutics. This is partly due to a wide variety in assay techniques to detect antibody responses to therapeutic proteins. Therefore, in this review we will discuss types of assays used for the detection of ADA and their (dis)advantages including factors influencing the results. Recent findings on clinical implications of immunogenicity will be discussed thereafter, and attention will be paid to ways to minimize the effect of immunogenicity. Finally, published algorithms on how to deal with immunogenicity in clinical practice will be reviewed.

ASSAY TECHNIQUES

Various classes of immunoglobulins, with variable affinity, can be produced. Furthermore, the formation of ADA fluctuates over time and immunological tolerance develops at a certain time point during treatment. In addition, timing of blood sampling during treatment can affect the detection of immunogenicity.

F1-10
Box 1:
no caption available

Several assay formats are available for the detection of ADA [2,3▪▪]. None of the available assays is able to detect all of the subclasses or idiotypes of ADA. IgM antibodies have low avidity, and are therefore difficult to detect. Affinity maturated IgG1 and IgG4 antibodies are easier to detect. IgE can also have high avidity, but excess IgG makes detection of the relatively low amounts of IgE challenging.

Commonly used assays include direct enzyme linked immunosorbent assays (ELISAs), which are straightforward but suffer from nonspecific binding [2,3▪▪]. Two-site or bridging assays are sensitive and specific, but do not detect IgG4 antibodies and are susceptible to drug interference [2,3▪▪]. Antigen binding tests (ABT) have been shown to be specific and are less susceptible to drug interference. However, for these radioimmunoassays there is a need for laboratory facilities for radioactivity [2,3▪▪]. A fourth type are cell-based assays which are not used frequently for immunogenicity testing, because they are time consuming, complex and variable [4].

CONFOUNDING FACTORS: FALSE POSITIVE

Numerous factors could lead to false-positive results when assessing the immunogenicity of therapeutic antibodies. Binding to the plate can induce conformational changes of the therapeutic protein that may result in nonspecific signals. In particular, Fc interactions of IgG4 with coated IgG may give rise to background problems [5–7]. In patients with rheumatoid arthritis (RA), low avidity IgM (or IgG) rheumatoid factor can also lead to false-positive results, because rheumatoid factors bind to the Fc part of antibodies [8]. Therefore, the use of F(ab)2 fragments of therapeutic antibodies in assays can eliminate this background, but it introduces another source of false-positive results due to antihinge antibodies (found in sera from some healthy individuals as well as some patients with rheumatoid arthritis). However, this background can be blocked by addition of polyclonal F(ab)2 fragments prepared from intravenous immunoglobulin [9].

CONFOUNDING FACTORS: FALSE NEGATIVE

Although the two-site assay is the most sensitive for the detection of ADA, a smaller number of patients test positive in the two-site assay compared with the ABT [3▪▪]. This discrepancy is caused by the susceptibility for drug interference. The two-site assay typically measures ADA in the absence of detectable drug levels. In the presence of drug, drug-antidrug antibody complexes are formed and as a result, the assay cannot detect the ADA. The ABT is able to measure a small part of the ADA in complex with the drug. Recently developed pH-shift-antiidiotype ABTs enabled the measurement of ADA in the presence of the drug by dissociation of the complexes [10,11]. Using this method, it was discovered that a large proportion of patients developed ADA; however, the majority of patients who developed ADA did not make sufficient quantities to neutralize all of the drug. These studies highlight the importance of the quantity of antibodies that are produced, both in assessment of immunogenicity as well as in the interpretation of the clinical consequences.

CLINICAL IMPLICATIONS: ADVERSE EVENTS

Immunogenicity has clinical implications for treatment with biological therapeutics. The presence of ADA is associated with adverse events, of which infusion reactions are the most common.

A well known example of an adverse reaction that occurred after the administration of an erythropoietin product was pure red cell aplasia [12], in which ADA against the erythropoietin product depleted not only the drug itself but also the endogenous erythropoietin. This very rare phenomenon (191 cases world wide between 1998 and 2003) was caused by conformational changes of the ADA and is typical for endogenous therapeutic proteins with a biological function, with potentially severe consequences for patients.

Infusion reactions are common adverse events in patients treated with biological medications, with limited consequences in some cases and major ones in others. There is a variable clinical presentation, ranging from mostly mild, for example slight decrease in blood pressure or erythema, to anaphylactic reactions necessitating stopping the infusion. The immunological reaction is immune complex (IgG) or IgE (allergic hypersensitivity reactions) related and could be dependent on the size of the immune complexes formed. For example, in infliximab-treated Crohn's disease patients, the presence of higher concentrations of ADA predicted a higher risk of infusion reactions (relative risk 2.40; 95% confidence interval 1.65–3.66) [13]. In another study [14], two nonresponding and two responding infliximab-treated RA patients received an infusion with radio labelled infliximab. Various sizes of infliximab–antiinfliximab complexes, including very large ones, were present in the serum of one of the nonresponding patients who had developed a severe infusion reaction. The other nonresponding patients had only small complexes present. A faster clearance of infliximab was found in this nonresponding patient and uptake in liver and spleen was higher compared with the responding patients.

In patients treated with cetuximab, a chimeric monoclonal antibody against epidermal growth factor receptor (approved for colorectal cancer and squamous cell carcinoma), a higher prevalence of hypersensitivity reactions was reported in some areas of the United States [15]. In most patients experiencing a hypersensitivity reaction IgE ADA were already present in the serum before therapy. These antibodies were directed against galactose-α-1,3-galactose (a-gal) which is present on the F(ab) fragment of cetuximab. A-gal is present on proteins of nonprimate animals and production of these IgE antibodies might be stimulated by bites from ticks. In addition, a-gal can be present as part of the conserved glycan in the Fc part of other antibodies such as infliximab. However, anti-a-gal antibodies were found not to bind to this epitope [16].

Adverse events associated with the formation of drug–antidrug complexes are probably rare. In some adalimumab-treated patients who previously had developed ADA, serious thromboembolic events occurred [17]. No firm conclusions can be drawn because this study was underpowered and other risk factors could have contributed to the occurrence of events. However, awareness is raised for the potential harm of ADA and immune complexes.

CLINICAL IMPLICATIONS: EFFICACY

Therapeutic antibodies are removed from the circulation at a similar or slightly faster rate compared with endogenous IgG. The production of ADA can cause faster clearance due to the formation of immune complexes. These complexes are taken up by liver and spleen, and therefore enhance clearance. In addition, drug in complex with ADA will not be biologically active if the ADAs are directed to the antigen-binding site, which is usually the only foreign part of humanized or fully human therapeutic antibodies. These two mechanisms lead to decreased drug levels and consequently to impaired treatment responses.

Immunogenicity in clinical practice is registered in several long-term observational cohort studies. The immunogenicity testing strategies are, in general, well defined and give insight into the course of the immune response against biological therapeutics and its clinical relevance.

In a Spanish study [18▪], 85 infliximab-treated RA patients were followed for more than 4 years. The formation of ADA (32.9%) was associated with poor clinical response: all European League Against Rheumatism (EULAR) nonresponders had developed ADA and ADA levels were higher in these patients compared with responding patients with ADA. Patients with ADA discontinued treatment more often compared with patients without ADA (82.1 vs. 39.3%) and the median survival time was shorter.

Similar results were found in a study [19] with 17 RA and 91 spondyloarthritis patients. Infliximab maintenance was lower in patients who developed ADA. In addition, trough infliximab concentration during the initiation phase of treatment was lower in patients who developed ADA during follow-up.

In 27 infliximab treated RA patients the course of infliximab and ADA levels were assessed within one infusion cycle [20▪]. One-fifth of the patients had nontherapeutical infliximab levels halfway through the infusion cycle. ADA was detected at that time point in most patients. These findings might have implications for the clinical response and perhaps radiologic progression in the long term.

In 272 adalimumab treated RA patients with 3 years of follow-up, development of ADA (28%) was associated with treatment discontinuation due to failure [21▪▪]. In addition, sustained minimal disease activity or sustained remission was less often achieved.

In contrast, in 292 etanercept treated RA patients, no neutralizing ADAs were detected during 6 months’ follow-up [22▪]. After 6 months of therapy etanercept levels were significantly higher in EULAR good responders (median 3.78 mg/l) compared with both moderate (3.10 mg/l) and nonresponders (2.80 mg/l). Forty percent of all nonresponding patients had an etanercept level below 2.1 mg/l, which is considerably lower than the average concentration of 3 mg/l found in pharmacokinetic studies.

From these studies, one could conclude that the presence of functional drug levels is required for an adequate treatment response. Therefore, clinical practice should not only be aimed at clinical response but also at adequate drug levels.

FACTORS INFLUENCING IMMUNOGENICITY

In RA patients treated with biologicals, efficacy is in general increased when prescribed in combination with methotrexate. Whether this effect is synergistic or related to suppression of immunogenicity is unresolved. In infliximab-treated patients taking concomitant methotrexate, infliximab trough levels were higher compared with patients treated with infliximab monotherapy [23]. There appears to be a favourable effect of cotreatment with methotrexate on the immunogenicity of biologicals [24]. The frequency and amount of ADA developed is reduced in patients concomitantly taking methotrexate [13,24,25].

Suppression of the early T-cell and B-cell expansion by methotrexate might be responsible for the modulation of the immune response, whereby the formation of ADA is reduced in quantity and subsequently drug levels are increased.

Others hypothesize that infliximab trough levels are higher in patients concomitantly treated with methotrexate because methotrexate reduces inflammation and consequently the tumour necrosis factor (TNF) load is lower, resulting in higher infliximab levels. Infliximab treated RA patients with high baseline C-reactive protein (CRP) levels (indicating active inflammation and high TNF levels) showed lower infliximab trough levels during the first 3 months of therapy [25].

Patients with high baseline disease activity had higher levels of ADA after 3 months of therapy compared with patients with lower disease activity [25]. Adalimumab-treated patients who developed ADA during follow-up had more longstanding and more severe disease at baseline, more often erosive disease, higher disease activity score, CRP and erythrocyte sedimentation rate (ESR) [21▪▪]. This might indicate that the state of the immune system, an active inflammatory state, influences the immune response against the ‘non self’ therapeutic antibodies.

A higher frequency of ADA was found in adalimumab treated RA patients who carried the same IgG allotype as present on adalimumab compared with patients whose IgG allotype differed [26]. Allotypes represent slight differences in the amino acid sequences of the constant chains of IgG molecules. In particular IgG (Gm) allotypes are racially distributed and could be immunogenic for individuals. Patients carrying the G1m17-allotype might be more prone to ADA responses.

Polymorphisms of genes might also be associated with the formation of ADA. Distinct IL 10 polymorphisms were associated with the formation of ADA in adalimumab treated patients [27]. However, as of yet there have been no sufficiently powered studies to detect genes influencing immunogenicity.

IMPLEMENTATION OF IMMUNOGENICITY

Immunogenicity is an important feature in the treatment of biologicals; however, the knowledge gained over the past years is not or insufficiently applied to daily clinical practice.

As low serum drug levels are associated with lack or loss of clinical response, therapeutic drug monitoring (TDM) would be an important tool in clinical decision-making in patients with rheumatic diseases. Obstacles to be overcome before TDM can be implemented in daily practice include assay standardization and validation, consensus on the interpretation of serum drug concentrations and ADA assay values and an evidence based treatment algorithm and the validation thereof for clinical practice [28].

Patient groups and time point(s) for TDM of biologicals have to be established. It has to be investigated, for example, whether we should measure serum from all biologics-treated patients (nonresponders as well as responders) at a predefined time point, for example 3 or 6 months, or just patients with lack or loss of response at any time during treatment.

Routine determination of drug levels could aid in elucidating determinants for (non)-response and could aid in clinical decision-making because the presence or absence of ADA has implications for the response to a second TNF-inhibitor [29,30,31▪▪]. In patients without ADA to their first TNF-inhibitor, a second TNF-inhibitor is less likely to be effective. These patients might benefit from biologicals with another mode of action. In contrast, patients who did develop ADA to their first TNF-inhibitor may respond to a second TNF-inhibitor comparably to anti-TNF naive patients. However, they are more prone to develop ADA to their second TNF-inhibitor.

Trough levels can indicate whether the patient is a primary (no response despite adequate drug levels) or secondary (no response due to inadequate drug levels) nonresponder. Important for definitions of lack or loss of response is the awareness that ADA are already present during the first weeks of treatment in a proportion of the patients [21▪▪].

Several algorithms for TDM are proposed for mainly RA and Crohn's disease [31▪▪,32▪,33]. These algorithms focus on patients with loss of response. Treatment strategies include shortening of the dose interval, dose escalation, switching to a second TNF-inhibitor or switching to a biological with another mechanism of action. The role of dose increase or shortening of dose interval remains unclear because measuring drug levels is usually not included in previously conducted studies [34]. In observational studies, after dose escalation, ADA titers decreased, without restoration of the clinical response [21▪▪]. On the contrary, in patients without ADA but with low-drug levels, dose increase resulted in restoration of clinical response and median-survival time was comparable with patients with adequate drug levels without the need for dose increase [31▪▪].

TDM could lead to a more nuanced approach and this should be investigated further. It is of interest that therapeutic alternatives vary between inflammatory diseases, and therefore separate treatment algorithms may be needed for each distinct inflammatory disease, as switching to a biological with another mechanism of action is, until now, not an available option for all diseases.

Some argue to measure drug levels (and ADA) in responding patients to determine the reason for response: despite inadequate drug levels or due to adequate drug levels [32▪]. In this first group of patients, discontinuation of expensive drugs could be considered as an option, because it is questionable whether low-disease activity can be attributed to the drug (neutralized by ADA). Development of dose-response curves could aid in rationally decreasing the dose of biologicals.

CONCLUSION

Several widely used biologicals possess large variation in pharmacokinetics, and therefore measuring drug levels seems appropriate. Measurement of drug levels is straightforward and the costs are low. ADA assays face confounding factors and results differ between types of assays potentially leading to confusion or misinterpretation when the user is unaware of these aspects.

Immunogenicity has a large and clinically relevant negative impact on the efficacy of biological therapeutics. ADA neutralize drug levels, and therefore clinical response is impaired.

Prevention of ADA formation is not yet feasible; however, concomitant use of methotrexate lowers the impact of immunogenicity on therapy. It should be investigated further whether the efficacy of TNF inhibitors in patients with, for example spondylarthropathy can be increased by concomitantly prescribing methotrexate, a drug not effective by itself in the treatment of spondylarthropathies.

Algorithms are necessary to adopt treatment strategies in the clinical setting in case of loss or lack of response and to treat patients as optimally as possible.

In summary, therapeutic drug monitoring will lead to more rational clinical decision-making that is based not only on clinical outcomes but also on pharmacokinetic parameters.

Acknowledgements

None.

Conflicts of interest

C.K., none declared; T.R., received payment for lectures from Pfizer and research support from Genmab; G.W., has received a research grant from Pfizer (Wyeth) and payment for lectures form Pfizer and Amgen.

REFERENCES AND RECOMMENDED READING

Papers of particular interest, published within the annual period of review, have been highlighted as:

  • ▪ of special interest
  • ▪▪ of outstanding interest

Additional references related to this topic can also be found in the Current World Literature section in this issue (p. 346).

REFERENCES

1. Division of Communication and Consumer AffairsOffice of Communication OaDFaDA. What is a biological product? http://www.hhs.gov/fda/faq/faq/1612.html. [Accessed 30 November 2010]
2. Wolbink GJ, Aarden LA, Dijkmans BA. Dealing with immunogenicity of biologicals: assessment and clinical relevance. Curr Opin Rheumatol 2009; 21:211–215.
Hart MH, de Vrieze H, Wouters D, et al. Differential effect of drug interference in immunogenicity assays. J Immunol Methods 2011; 372 (1–2):196–203.

Drug interference is one of the potential confounders in the assessment of immunogenicity. Level of susceptibility differs between assay types. Knowledge of drug interference helps in the interpretation of studies on immunogenicity.

4. Wadhwa M, Thorpe R. Unwanted immunogenicity: lessons learned and future challenges. Bioanalysis 2010; 2:1073–1084.
5. Rispens T, Ooievaar-De Heer P, Vermeulen E, et al. Human IgG4 binds to IgG4 and conformationally altered IgG1 via Fc-Fc interactions. J Immunol 2009; 182:4275–4281.
6. Rispens T, Leeuwen A, Vennegoor A, et al. Measurement of serum levels of natalizumab, an immunoglobulin G4 therapeutic monoclonal antibody. Anal Biochem 2011; 411:271–276.
7. Zack DJ, Stempniak M, Wong AL, Weisbart RH. Localization of an Fc-binding reactivity to the constant region of human IgG4. Implications for the pathogenesis of rheumatoid arthritis. J Immunol 1995; 155:5057–5063.
8. Tatarewicz S, Miller JM, Swanson SJ, Moxness MS. Rheumatoid factor interference in immunogenicity assays for human monoclonal antibody therapeutics. J Immunol Methods 2010; 357:10–16.
9. Rispens T, de Vrieze H, de Groot E, et al. Antibodies to constant domains of therapeutic monoclonal antibodies: antihinge antibodies in immunogenicity testing. J Immunol Methods 2011; 375:93–99.
10. van Schouwenburg PA, Bartelds GM, Hart MH, et al. A novel method for the detection of antibodies to adalimumab in the presence of drug reveals ‘hidden’ immunogenicity in rheumatoid arthritis patients. J Immunol Methods 2010; 362 (1–2):82–88.
11. Lofgren JA, Wala I, Koren E, et al. Detection of neutralizing antitherapeutic protein antibodies in serum or plasma samples containing high levels of the therapeutic protein. J Immunol Methods 2006; 308 (1–2):101–108.
12. Bennett CL, Luminari S, Nissenson AR, et al. Pure red-cell aplasia and epoetin therapy. N Engl J Med 2004; 351:1403–1408.
13. Baert F, Noman M, Vermeire S, et al. Influence of immunogenicity on the long-term efficacy of infliximab in Crohn's disease. N Engl J Med 2003; 348:601–608.
14. van der Laken CJ, Voskuyl AE, Roos JC, et al. Imaging and serum analysis of immune complex formation of radiolabelled infliximab and antiinfliximab in responders and nonresponders to therapy for rheumatoid arthritis. Ann Rheum Dis 2007; 66:253–256.
15. Chung CH, Mirakhur B, Chan E, et al. Cetuximab-induced anaphylaxis and IgE specific for galactose-alpha-1,3-galactose. N Engl J Med 2008; 358:1109–1117.
16. van Bueren JJ, Rispens T, Verploegen S, et al. Antigalactose-α-1,3-galactose IgE from allergic patients does not bind α-galactosylated glycans on intact therapeutic antibody Fc domains. Biotechnol 2011; 29:574–576.doi: 10.1038/nbt.1912.
17. Korswagen LA, Bartelds GM, Krieckaert CL, et al. Venous and arterial thromboembolic events in adalimumab-treated patients with antiadalimumab antibodies: a case series and cohort study. Arthritis Rheum 2011; 63:877–883.doi: 10.1002/art.30209.
Pascual-Salcedo D, Plasencia C, Ramiro S, et al. Influence of immunogenicity on the efficacy of long-term treatment with infliximab in rheumatoid arthritis. Rheumatology (Oxford) 2011; 50:1445–1452.

This study is the first to provide 10-year follow-up data on immunogenicity in an observational setting.

19. Ducourau E, Mulleman D, Paintaud G, et al. Antibodies toward infliximab are associated with low infliximab concentration at treatment initiation and poor infliximab maintenance in rheumatic diseases. Arthritis Res Ther 2011; 13:R105.
van den Bemt BJ, den Broeder AA, Wolbink GJ, et al. Antiinfliximab antibodies are already detectable in most patients with rheumatoid arthritis halfway through an infusioncycle: an open-label pharmacokinetic cohort study. BMC Musculoskelet Disord 2011; 12:12.

This study gives insight into the course of infliximab levels in ADA during one infusion cycle. Part of the patients are exposed to nontherapeutical drug levels already half way through the infusion cycle.

Bartelds GM, Krieckaert CL, Nurmohamed MT, et al. Development of antidrug antibodies against adalimumab and association with disease activity and treatment failure during long-term follow-up. JAMA 2011; 305:1460–1468.

In adalimumab treated RA patients, immunogenicity impaired important clinical treatment outcomes. ADA developing patients less often achieved, for example, remission, shedding light on the impact of immunogenicity.

Jamnitski A, Krieckaert CL, Nurmohamed MT, et al. Patients nonresponding to etanercept obtain lower etanercept concentrations compared with responding patients. Ann Rheum Dis 2012; 71:88–91.

Even for biologicals that appear to be less immunogenic (no neutralizing ADA) it seems appropriate to measure drug levels.

23. Vermeire S, Noman M, Van Assche G, et al. Effectiveness of concomitant immunosuppressive therapy in suppressing the formation of antibodies to infliximab in Crohn's disease. Gut 2007; 56:1226–1231.
24. Krieckaert CL, Bartelds GM, Lems WF, Wolbink GJ. The effect of immunomodulators on the immunogenicity of TNF-blocking therapeutic monoclonal antibodies: a review. Arthritis Res Ther 2010; 12:217.
25. Bendtzen K, Geborek P, Svenson M, et al. Individualized monitoring of drug bioavailability and immunogenicity in rheumatoid arthritis patients treated with the tumor necrosis factor alpha inhibitor infliximab. Arthritis Rheum 2006; 54:3782–3789.
26. Bartelds GM, de Groot E, Nurmohamed MT, et al. Surprising negative association between IgG1 allotype disparity and antiadalimumab formation: a cohort study. Arthritis Res Ther 2010; 12:R221.
27. Bartelds GM, Wijbrandts CA, Nurmohamed MT, et al. Antiadalimumab antibodies in rheumatoid arthritis patients are associated with interleukin-10 gene polymorphisms. Arthritis Rheum 2009; 60:2541–2542.
28. Colombel JF, Feagan BG, Sandborn WJ, et al. Therapeutic drug monitoring of biologics for inflammatory bowel disease. Inflamm Bowel Dis 2012; 18:349–358.
29. Jamnitski A, Bartelds GM, Nurmohamed MT, et al. The presence or absence of antibodies to infliximab or adalimumab determines the outcome of switching to etanercept. Ann Rheum Dis 2011; 70:284–288.
30. Bartelds GM, Wijbrandts CA, Nurmohamed MT, et al. Antiinfliximab and antiadalimumab antibodies in relation to response to adalimumab in infliximab switchers and antitumour necrosis factor naive patients: a cohort study. Ann Rheum Dis 2010; 69:817–821.
Afif W, Loftus EV Jr, Faubion WA, et al. Clinical utility of measuring infliximab and human antichimeric antibody concentrations in patients with inflammatory bowel disease. Am J Gastroenterol 2010; 105:1133–1139.

Algorithm for biological treated Crohn's disease patients and an overview of the evidence from literature for each proposed option. Although for Crohn's disease patients, also applicable for RA patients.

Bendtzen K. Is there a need for immunopharmacologic guidance of antitumor necrosis factor therapies? Arthritis Rheum 2011; 63:867–870.doi: 10.1002/art.30207.

Proposed algorithm for biological treated RA patients. Drug levels and ADA, according to this algorithm, should have implications for responding as well as nonresponding patients.

33. Yanai H, Hanauer SB. Assessing response and loss of response to biological therapies in IBD. Am J Gastroenterol 2011; 106:685–698.
34. Blom M, Kievit W, Kuper HH, et al. Frequency and effectiveness of dose increase of adalimumab, etanercept, and infliximab in daily clinical practice. Arthritis Care Res (Hoboken) 2010; 62:1335–1341.
Keywords:

assay techniques; biological therapies; clinical implications; immunogenicity; therapeutic drug monitoring

© 2012 Lippincott Williams & Wilkins, Inc.