Therapeutic drug monitoring for infliximab & adalimumab in inflammatory bowel disease
Tumor necrosis factor-alpha (TNF-α) is a master pro-inflammatory cytokine that drives the chronic mucosal inflammation seen in Crohn's disease and ulcerative colitis. Anti-TNF biologics — infliximab and adalimumab foremost among them — neutralize TNF-α directly at the site of disease, but their real-world effectiveness depends critically on maintaining an adequate, sustained drug exposure between doses.
TNF-α is produced predominantly by activated lamina propria macrophages, dendritic cells, and T cells in the inflamed intestinal wall. It exists in two biologically active forms: a 26 kDa transmembrane precursor (tmTNF) that signals cell-to-cell, and a soluble 17 kDa trimer (sTNF) released after cleavage by the metalloprotease TACE (ADAM17), which acts at a distance through the circulation and interstitium.
Both forms signal through TNFR1 (p55, expressed nearly ubiquitously, drives apoptosis and inflammatory gene transcription via NF-κB) and TNFR2 (p75, restricted to immune and endothelial cells, drives proliferative and survival signals). In IBD, excess TNF-α upregulates adhesion molecules on gut endothelium (ICAM-1, VCAM-1), recruits neutrophils and monocytes, stimulates matrix metalloproteinases that erode the epithelial barrier, and sustains a self-amplifying Th1/Th17 inflammatory loop.
Mucosal TNF-α concentrations correlate with endoscopic severity, and TNF-α is detectable in stool and serum of active IBD patients at levels that fall with successful treatment — making it both a driver of disease and a biomarker of response.
Anti-TNF agents were the first biologic class approved for IBD — infliximab in 1998 for fistulizing Crohn's disease — and remain the most widely used and best-studied biologic class in gastroenterology, with over two decades of real-world outcome and immunogenicity data.
Infliximab and adalimumab are both bivalent IgG1 monoclonal antibodies with high affinity (sub-nanomolar KD) for the TNF-α trimer. Infliximab is a chimeric antibody — murine variable (antigen-binding) regions grafted onto a human IgG1 constant region — while adalimumab is fully human, produced by phage-display selection against TNF-α, which reduces (but does not eliminate) its immunogenic potential.
Mechanistically, anti-TNF antibodies act through several complementary pathways:
• Direct neutralization: binding soluble TNF-α trimers prevents them from engaging TNFR1/TNFR2 on target cells, blocking downstream NF-κB and MAPK inflammatory signaling. • Transmembrane TNF binding: binding tmTNF on activated macrophages and T cells triggers "reverse signaling" back into the TNF-bearing cell, along with antibody-dependent cellular cytotoxicity (ADCC) and complement-dependent cytotoxicity (CDC) that eliminates the inflammatory cell itself — a mechanism thought to be particularly important for mucosal healing and one that differs somewhat between infliximab and adalimumab. • Induction of apoptosis: engagement of tmTNF-bearing lamina propria T cells and monocytes promotes their apoptotic clearance, directly reducing the pool of TNF-producing cells in the gut wall. • Immune complex formation: the IgG1 Fc region also allows opsonization and clearance of TNF-antibody complexes by the reticuloendothelial system.
Because anti-TNF neutralization is stoichiometric and concentration-dependent, clinical efficacy requires that free drug remains present in the tissue and circulation throughout the entire dosing interval — not just immediately after infusion. If serum drug concentration falls too low before the next dose, TNF-α escapes neutralization, inflammation resumes, and symptoms recur ("loss of response").
This single pharmacological fact — that efficacy tracks drug exposure, not just drug identity — is the entire rationale for therapeutic drug monitoring (TDM). Roughly 10–20% of patients fail to respond to anti-TNF induction therapy (primary non-response), and an additional 23–46% of initial responders lose response over the subsequent 1–3 years of maintenance therapy (secondary loss of response), most commonly because trough levels have fallen below the threshold needed to fully neutralize TNF-α at the tissue level.
Every infused or injected dose of a biologic produces a characteristic pharmacokinetic curve: a rapid rise to peak concentration, followed by a slower biexponential decline until the next dose is due. The lowest point of that curve — the trough — is the pharmacologically decisive value, because it represents the drug exposure the gut tissue experiences for the longest stretch of the interval.
After an IV infusion of infliximab, serum concentration follows a two-compartment pharmacokinetic model: a fast initial distribution phase (drug equilibrates from plasma into extravascular and gut tissue compartments over hours to a few days) followed by a slower terminal elimination phase governed by catabolism through the reticuloendothelial (Fc receptor-mediated) system, with a terminal half-life of roughly 7.7–9.5 days in patients without anti-drug antibodies.
Adalimumab, given subcutaneously, shows a slower absorption phase (peak serum levels at ~5 days post-injection) but a longer terminal half-life of 10–20 days (label average ≈ 14 days), reflecting both its fully human Fc structure and its SC depot-release kinetics.
The "trough" concentration (Ctrough) is simply the drug level at the very end of the dosing interval — immediately before the next infusion or injection — and represents the pharmacokinetic nadir the tissue is exposed to. Because the interval between doses is fixed by regimen (e.g., every 8 weeks for infliximab, every 2 weeks for adalimumab), the trough is mathematically determined by the peak concentration, the elimination rate constant, and the interval length: Ctrough = C0 · e^(−k·t), where k = ln(2)/t½.
A shorter half-life or a longer interval both push the trough down toward — or below — the therapeutic floor. This is why the two most direct pharmacokinetic levers for a falling trough are exactly the two the clinician controls: raise the dose (raises C0) or shorten the interval (reduces the time available for decay).
Unlike small-molecule drugs, monoclonal antibody clearance varies substantially between patients due to factors largely unrelated to renal or hepatic function:
• Body size and albumin: larger body surface area and low serum albumin (a marker of disease severity and gut protein loss) both accelerate clearance — infliximab is partly lost through an inflamed, leaky gut mucosa in severe colitis, a phenomenon sometimes called "fecal loss" of drug. • TNF antigen burden: very high inflammatory burden consumes more drug through target-mediated drug disposition, transiently accelerating apparent clearance during active flares. • Concomitant immunomodulators: co-administration of azathioprine, 6-mercaptopurine, or methotrexate reduces immunogenicity and stabilizes clearance (this was the central finding of the SONIC trial, Colombel et al. 2010). • Anti-drug antibodies (ADA): the single largest driver of accelerated, unpredictable clearance — see below. • Sex, smoking status, and disease phenotype contribute smaller but measurable effects on trough variability across cohorts.
When the immune system mounts a response against the therapeutic antibody itself — anti-drug antibodies (ADA), historically called antibodies-to-infliximab (ATI) — the resulting drug-ADA immune complexes are rapidly cleared by the reticuloendothelial system, dramatically shortening the effective half-life. High-titer, persistent ADA can reduce the functional half-life of infliximab from roughly 9–10 days to well under 5 days, causing the trough to collapse even when the nominal dose and interval are unchanged.
ADA formation is itself favored by low trough levels (periods where circulating drug is scarce give the immune system an opportunity to recognize the antibody as foreign), creating a vicious cycle: low trough → ADA formation → faster clearance → even lower trough → treatment failure. This feedback loop is the central pharmacological argument for proactive monitoring — intervening before the cycle becomes self-sustaining.
Therapeutic drug monitoring requires a blood sample drawn at the correct moment — immediately before the next scheduled dose — and a laboratory assay capable of separately quantifying free drug concentration and anti-drug antibody titer, even when both are present in the same sample simultaneously.
The most widely used assay format is a sandwich ELISA: a microtiter plate is coated with TNF-α (or an anti-idiotype antibody), the patient's serum sample is added, and any free (drug-tolerant assays even detect drug still bound to some ADA) infliximab or adalimumab in the sample binds the coated TNF-α via its antigen-binding site. A secondary enzyme-labeled anti-human-IgG antibody then binds the captured drug, and a colorimetric substrate reaction produces a signal proportional to drug concentration, read against a calibration curve of known standards.
ADA is measured in a mirror-image format: the plate is coated with the drug itself (infliximab or adalimumab), patient serum is added, and any anti-drug antibodies present bind the coated drug; a labeled anti-human-IgG secondary detects them. Older ("drug-sensitive") assays could not detect ADA in the presence of any free drug because bridging was blocked — a major limitation, since drug and ADA often coexist. Modern "drug-tolerant" assays (acid dissociation, or newer methods like homogeneous mobility shift assay [HMSA] and reporter-gene bioassays) dissociate drug-ADA complexes before measurement, allowing accurate ADA titers even in patients with detectable trough levels.
Because different laboratories use different assay platforms and calibrators, absolute trough and ADA values are not always perfectly interchangeable between labs — clinicians are advised to use the same assay consistently for serial monitoring in an individual patient.
Two distinct clinical strategies use the same assay technology for different purposes:
• Reactive TDM: testing is triggered by loss of response — recurrent symptoms, rising CRP or fecal calprotectin, or endoscopic relapse — to determine whether the problem is pharmacokinetic (low trough), immunogenic (high ADA), or mechanistic (adequate trough, active disease). This is the more established, guideline-endorsed use of TDM and has high-quality evidence supporting its cost-effectiveness and ability to guide dose changes versus empiric drug switching.
• Proactive TDM: trough levels are checked on a fixed schedule (e.g., at the end of induction around week 14, and periodically during maintenance) regardless of symptoms, with dose adjusted to maintain a target trough before loss of response ever becomes clinically apparent. The rationale is the ADA feedback loop described in Stage 2 — intervening while trough is falling but before antibodies become high-titer and clearance becomes unrecoverable.
Professional society guidance (e.g., American Gastroenterological Association 2021 clinical practice update) supports reactive TDM as standard of care and endorses proactive TDM as a reasonable strategy, particularly around the end of induction, though its incremental benefit over well-conducted reactive testing remains an area of active study.
Correct sample timing is essential: a level drawn even a few days before the true trough (next-dose date) can substantially overestimate the nadir concentration because of ongoing biexponential decay. Point-of-care and rapid lab-based assays (turnaround as fast as same-day) increasingly allow "treat-to-target" decisions to be made at the same visit as the infusion, avoiding a delay between blood draw and dose adjustment.
Most commercial and hospital-lab assay panels report free drug concentration in micrograms per milliliter (μg/mL) and ADA as a titer (arbitrary units per mL, AU/mL, or as a categorical negative/low-positive/high-positive/positive-with-detectable-drug result), which clinicians then interpret against published therapeutic thresholds.
A single trough level and ADA titer, plotted together, sort every case of anti-TNF non-response into one of four clinically distinct categories — each demanding a different next step. This quadrant framework, popularized by Roda et al. and refined across multiple TDM cohorts, is the conceptual core of modern anti-TNF management in IBD.
A subtherapeutic trough with negative or low-titer ADA indicates the drug is simply being cleared too quickly relative to the dose and interval — an underdosing problem, not an immune problem. This is the most actionable and favorable scenario for dose optimization: increasing the dose or shortening the interval typically restores an adequate trough and clinical response, because there is no antibody actively destroying the drug.
This pattern is common in patients with high inflammatory burden (accelerated target-mediated clearance), low albumin, high body weight, or simply standard dosing that was never intensive enough for that individual's clearance rate.
A subtherapeutic trough accompanied by a high ADA titer indicates the immune system has mounted a durable response against the drug itself, accelerating its clearance faster than any feasible dose increase can compensate for. Dose escalation in this setting is usually futile — the added drug is simply neutralized and cleared by the same antibody response.
The standard recommendation here is to switch to a different anti-TNF agent (within-class switch, e.g., infliximab to adalimumab) — because ADA are typically drug-specific and do not cross-react between different anti-TNF molecules — ideally paired with a concomitant immunomodulator to reduce the risk of forming new antibodies against the second agent. High and sustained ADA titers, especially with undetectable drug on repeated testing, are associated with the lowest probability of ever recapturing response to the same drug at any dose.
When the trough is comfortably within — or even above — the target therapeutic band, and ADA are negative, but the patient still has objective evidence of active inflammation (elevated fecal calprotectin, endoscopic ulceration), the drug is present in adequate amounts but is simply not achieving disease control. This is a true mechanistic failure of the TNF-blockade strategy itself, not a dosing problem.
Further dose escalation offers little additional benefit in this quadrant. The appropriate step is switching out of class to a biologic or small molecule with a different mechanism of action — for example vedolizumab (gut-selective anti-α4β7 integrin), ustekinumab (anti-IL-12/23 p40), risankizumab (anti-IL-23 p19), or a JAK inhibitor such as tofacitinib or upadacitinib — rather than cycling through another anti-TNF agent.
Adequate trough with negative or low ADA and quiescent disease (clinical remission, normal or near-normal fecal calprotectin, mucosal healing on endoscopy) is the goal state. The recommendation is to continue the current regimen, and — under a proactive monitoring strategy — periodically re-check trough levels to catch any future drift toward Quadrant A or B before symptoms recur, since pharmacokinetics can deteriorate gradually even in a patient currently doing well.
The same numeric trough value can sit in different quadrants for different patients depending on ADA status and clinical response — a level of "5 µg/mL" is not inherently good or bad; it must always be read together with the ADA titer and the clinical/biomarker picture.
The final step converts the quadrant classification into a concrete prescription: how much drug, how often, whether to add an immunomodulator, and whether to abandon the current agent altogether. Two landmark randomized trials — TAXIT and TAILORIX — directly tested whether trough-guided dosing algorithms outperform empiric, symptom-driven dosing.
For Quadrant A (pharmacokinetic failure), the standard infliximab escalation moves from 5 mg/kg every 8 weeks to either 10 mg/kg every 8 weeks or 5 mg/kg every 4–6 weeks — both approximately double the cumulative monthly dose and raise the trough proportionally, following the same first-order pharmacokinetic relationship (Ctrough = C0 · e^(−k·t)) used to model the decay curve in Stage 2. For adalimumab, escalation typically moves from 40 mg every 2 weeks to 40 mg weekly or 80 mg every 2 weeks.
Dose escalation is empirically supported: population pharmacokinetic modeling and real-world cohorts consistently show that raising trough into the 5–10 µg/mL band (or 10–15 µg/mL for fistulizing/complex Crohn's disease) recaptures clinical and endoscopic response in a substantial proportion of patients whose failure was pharmacokinetic rather than immunogenic or mechanistic.
For patients with low-titer, potentially reversible ADA — or as prevention in high-risk patients not yet immunized against the drug — adding azathioprine, 6-mercaptopurine, or methotrexate reduces both the rate of new ADA formation and, in some cases, the titer of existing antibodies, by suppressing the T-cell-dependent B-cell response against the therapeutic antibody. The SONIC trial (Colombel et al., NEJM 2010) demonstrated that infliximab plus azathioprine produced significantly higher steroid-free remission rates than infliximab monotherapy in Crohn's disease, an effect attributed substantially to reduced immunogenicity and higher, more stable trough concentrations in the combination arm.
Combination therapy carries its own trade-offs — increased risk of infection and, with thiopurines, lymphoma and non-melanoma skin cancer — so the decision to add or continue an immunomodulator must weigh immunogenicity benefit against these long-term safety considerations, especially in younger patients.
Quadrant B (immunogenic failure) calls for a within-class switch — moving to a different anti-TNF molecule, since ADA are typically specific to the original drug's idiotype and rarely cross-react. Within-class switching after immunogenic failure recaptures response in a majority of appropriately selected patients, particularly when paired with an immunomodulator to blunt antibody formation against the new agent.
Quadrant C (mechanistic failure) calls for an out-of-class switch to a biologic or small molecule with a distinct mechanism — vedolizumab, ustekinumab, risankizumab, or a JAK inhibitor — since the TNF pathway itself has been adequately blocked and inflammation persists through TNF-independent routes. Repeating anti-TNF exposure in this quadrant (cycling to a second anti-TNF agent) has a lower expected response rate than switching mechanism entirely.
The TAXIT trial (Vande Casteele et al., Gastroenterology 2015) randomized IBD patients on maintenance infliximab, first optimizing everyone to a target trough of 3–7 µg/mL, then comparing continued trough-based ("proactive") dosing against clinical symptom-based dosing. The initial optimization phase alone converted a large fraction of clinically stable-but-subtherapeutic patients to a more robust remission, and the proactive arm showed fewer undetectable troughs and flares during follow-up, even though the pre-specified primary remission endpoint at study end did not reach statistical significance between arms — an instructive result showing that TDM's main benefit may be preventing deterioration rather than producing a large one-time efficacy gain.
The TAILORIX trial (Colombel et al., Gastroenterology 2019) tested, in Crohn's disease patients starting infliximab, whether adding trough-level-based dose intensification on top of a symptom-and-biomarker-based algorithm improved sustained corticosteroid-free remission compared to the symptom/biomarker algorithm alone. It did not show a significant additional benefit of the trough-guided intensification, tempering enthusiasm for trough-only-driven escalation early in treatment and reinforcing that TDM should complement, not replace, clinical and biomarker assessment.
Taken together, the evidence supports TDM as a decision-support tool — most clearly valuable for interpreting loss of response and guiding rational (rather than empiric trial-and-error) treatment changes — rather than a fully automated dosing algorithm that alone dictates therapy.
| Product | Indication | Trial Design | Key Result |
|---|---|---|---|
| Pharmacokinetic failure | Low trough, ADA negative/low | Underdosing relative to individual clearance rate | Escalate dose or shorten interval |
| Immunogenic failure | Low trough, ADA high-titer | Antibody-accelerated clearance of drug | Switch within class ± add immunomodulator |
| Mechanistic failure | Adequate trough, active disease | TNF blockade adequate but insufficient to control inflammation | Switch out of class (different MOA) |
| Adequate therapy | Adequate trough, quiescent disease | Target achieved, disease controlled | Continue; monitor proactively over time |