Optimizing Infliximab Therapy for Inflammatory Bowel Disease— The Tools Are Getting Sharper

Infliximab (Remicade, Janssen Biotech) is a chimeric (75% human and 25% murine), monoclonal, immunoglobulin (Ig)G1 antibody that binds to soluble tumor necrosis factor (TNF)-α and prevents the cytokine from triggering the cellular TNF receptor complex and its effects. Infliximab also binds to transmembrane TNF-α and results in apoptosis of TNF-α-producing cells.¹ Up to 40% of Crohn's disease (CD) patients who initially respond to infliximab lose response within the first year.² Secondary nonresponse may be due to disease-related factors or drug-related factors, including neutralizing antibodies, altered clearance of the drug, or immunologic escape from TNF-driven inflammation. Recent guidelines from the World Congress of Gastroenterology suggest that a diminished or suboptimal response to infliximab can be managed in 1 of 3 ways: shortening the interval between doses, increasing the dose to 10 mg/kg, or switching to a different anti-TNF agent (in patients who continue to have loss of response after receiving the increased dose).³

Failure of infliximab therapy may be due to pharmacokinetic or pharmacodynamic mechanisms or immunogenic mechanisms. Serum albumin may be predictive of infliximab pharmacokinetics.⁴ All exogenous proteins have the potential to induce immunogenicity.⁵ The formation of anti-infliximab antibodies (ATIs) is associated with a lower serum infliximab level, diminished clinical response, and infusion reactions.⁶ In the SONIC study, ATIs were detected at Week 30 in 0.9% of patients receiving combination therapy with azathioprine plus infliximab and 14.6% of patients receiving infliximab monotherapy.⁷ Median serum trough levels of infliximab were higher in the combination therapy group than the infliximab monotherapy group.

The most commonly used method for detection of ATIs is a double-antigen enzyme-linked immunosorbent assay (ELISA) that uses specific antibodies for capture and detection.⁸ Serum infliximab interferes with ATI measurement in this method. Infliximab is an IgG construct containing κ light chains. An alternative ELISA using an anti-human λ chain antibody for ATI detection is less amenable to interference and may be able to detect ATIs in patients with detectable serum infliximab. The presence of ATIs and detectable serum infliximab by this method may be a harbinger of evolving loss of response.⁹ The immunogenic part of infliximab is the Fab fragment, but measuring ATIs is more useful than measuring antibodies against Fab(2) or Fab fragments.¹⁰

Solid-phase ELISAs have a risk of false-positive results due to nonspecific binding to immunoglobulins other than infliximab.¹¹ The use of fluid-phase radioimmunoassay (RIA) rather than solid-phase tests (RIA or ELISA) improves the specificity of the assay.¹² RIA is not influenced by artifacts induced by solid-phase adsorption of proteins. Fluid-phase RIA measures the functional bioactive infliximab concentration that is not neutralized by ATIs and therefore remains capable of neutralizing TNF-α. Fluid-phase RIA reports the TNF-α binding capacity expressed as infliximab equivalents μg/mL). ATIs (all isotypes) are detected when they bind to ¹²⁵ I-infliximab, after which they are separated by anti-human λ light chain antibodies.

A retrospective study published by Afif and colleagues in 2010 examined the utility of measuring ATIs and infliximab concentrations (by ELISA) in the management of inflammatory bowel disease patients.¹³ The authors found that increasing the infliximab dose in patients who have ATIs was ineffective, but increasing the dose in patients with subtherapeutic infliximab concentrations might be effective. Because the presence of infliximab in the sample interferes with the ATI assay, any patient with a detectable ATI concentration is considered by definition to have an undetectable infliximab concentration. Thus, 3 scenarios are possible: The patient can have a positive ATI test result; the patient can have a therapeutic infliximab concentration (defined as >12 mcg/mL at 4 weeks or a detectable trough level); or the patient can have a subtherapeutic infliximab concentration (defined as <12 mcg/mL at 4 weeks or an undetectable trough level). Afif and coauthors suggested a treatment algorithm for each situation, but interference in the ATI assay by infliximab limited the precision of interpretation.¹³ Reliable cutoff levels are necessary for both infliximab trough levels and ATI levels in order to anchor clinical decisions, but such cutoff levels were unavailable until recently.

In the current study by Steenholdt and colleagues, the authors attempted to determine clinically relevant cutoff values for infliximab trough levels and ATI levels associated with clinical response in patients with CD and ulcerative colitis (UC) by using fluid-phase RIA.¹⁴ Optimal cutoff levels to separate patients who maintained response from those who lost response were determined by using receiver operating characteristics analysis. The authors determined that a cutoff value of 0.5 μg/mL for infliximab trough level in CD patients provided a sensitivity of 86% and a specificity of 85%, with an accuracy of 87%. For UC patients, the cutoff level was 0.8 μg/mL, with a sensitivity of 75% and a specificity of 100%. The cutoff level for ATIs was 10 U/mL in both groups; this level corresponded to the detection limit of the assay. This level showed a sensitivity of 81% and a specificity of 90% in CD patients; in UC patients, this cutoff value yielded sensitivity and specificity values of 80% and 100%, respectively. The authors concluded that combining measurements of infliximab and ATIs had the highest overall accuracy (90%) in CD patients, with a sensitivity of 81% and a specificity of 94%. In this study, 20% of CD patients who lost response to infliximab had undetectable ATI levels; half of these patients had infliximab trough levels lower than the established level, while the other half had normal infliximab levels but still lost response. Unlike a previous Canadian study of UC patients, which showed a correlation between serum infliximab level and clinical response, the current study showed both low infliximab trough levels and high ATI levels in UC patients who lost response to maintenance infliximab therapy.^14,15

In terms of the study's limitations, the Steenholdt study was retrospective, and maintenance or loss of response was determined by chart reviews.¹⁴ The patient numbers were relatively small, especially for the UC group. The decision to continue or discontinue infliximab was based on clinical assessment by the gastroenterologist, not on infliximab trough level or ATI status. In addition, as in most studies, infliximab serum levels were measured as trough levels just prior to infliximab infusions but not at any other time point between infusions.

This study demonstrates that determination of clinically relevant, quantitative cutoff levels of infliximab and ATIs can be made with improved next-generation assays. In addition, the current study reinforces optimization of immunogenicity through the use of concomitant immunomodulators. Prospective studies are now required to base decision analysis on these cutoff levels and see whether they support intuitive treatment algorithms: increase in infliximab dosage (low infliximab trough levels, no ATIs), change to another anti-TNF monoclonal antibody (high ATI levels), or switch to another class of TNF inhibitors (adequate infliximab trough levels, no ATIs). A recent French study suggested that increasing the dose of infliximab may be effective irrespective of serum infliximab or ATI levels in patients who are losing response.¹⁶ This recommendation is current clinical practice in the majority of patients losing response to infliximab. Whether use of newer-generation assays with defined cutoff levels will provide better clinical decisions and outcomes will require prospective randomized trials, but at least the assays to conduct these trials optimally are becoming available.