Crohn's disease (CD) and ulcerative colitis (UC) are the two main types of chronic inflammatory bowel disease (IBD). Although their etiology remains unknown, studies have found an high concentration of tumor necrosis factor-alpha (TNF-α) in both CD and UC compared to controls1. In this regard, anti-TNF is an effective therapy to induce and maintain clinical remission as well as mucosa healing (MH).

Adalimumab (ADA) and infliximab (IFX) are the first-line anti-TNF drugs in case of inadequate response to conventional treatments in patients with moderately to severely active IBD. Although pharmacological options have increased in recent years with new biologic drugs (ustekinumab, vedolizumab…) and JAK inhibitors such as tofacitinib2, the available lines after failure of anti-TNF agents are limited.

However, about 10-30% of IBD patients treated with anti-TNF do not respond to induction therapy (primary non-response), while 23-46% of patients who are in remission lose response (LOR) during treatment (secondary LOR). The most common mechanism implicated in the development of LOR is the formation of antibodies against anti-TNF that interfere with TNF binding, accelerate drug clearance and correlate with lower serum drug concentrations and shorter duration of response3. The rate of LOR after 12 months ranges from 23% to 46% in CD patients treated with IFX or ADA4. Consequently, approximately 25% of CD patients experience LOR to IFX during the first year of maintenance therapy5. In addition, the pharmacokinetics of anti-TNFs are highly variable among patients and could be influenced by many factors, such as serum albumin, gender, body mass index (BMI), systemic inflammation and coadministration of immunosuppressive drugs (IMM)6.

Serum concentrations of these anti-TNFs have been correlated with improved clinical outcomes7. Since healthcare expenditure in IBD have been mainly attributed to drug treatment costs8, therapeutic drug monitoring (TDM) could be important in reducing healthcare costs without negatively impacting efficacy9. Deep remission (DR) defined as biomarker remission, clinical remission and mucosal healing has emerged as a new therapeutic target in IBD10. However, an optimal cutoff concentration of IFX and ADA predictive of DR has not been clearly defined.

The aim of this study was to define an optimal serum through concentrations cut-off point for IFX and ADA during maintenance therapy associated with DR in IBD. The secondary objective, was to evaluate the influence of sociodemographic, clinical, pharmacological and biomarker variables on anti-TNF concentrations and DR.

A treat-to-target approach leading to better clinical outcomes as a new therapeutic strategy. This study demonstrated that higher IFX and ADA concentrations during maintenance therapy are associated with DR in IBD.

TDM of anti-TNF agents has been associated with biomarker normalization and clinical remission. In a cohort based on a study of 1,487 samples from 483 patients, it was shown that an IFX concentration of > 2.79 µg/mL was associated with CRP < 5 mg/L12. A meta-analysis13, which included 22 studies, a higher minimum serum IFX concentration was associated with clinical remission, and an IFX cutoff concentration above 2 µg/mL was associated with an increased likelihood of clinical remission. Similarly, a major ADA concentrations were associated with improved clinical outcomes. The meta-analysis14 that included seven observational studies, a systematic review, a meta-analysis, and a clinical trial post hoc analysis, suggested ranges of ADA concentrations vary by target, postulating concentrations of 4.8 µg/mL as the cutoff point for clinical remission.

In our study, univariate and multivariate logistic regression analysis shows a strong correlation between IFX and ADA concentrations and DR. Importantly, an optimal cutoff value of 3.1 µg/mL for IFX and 6.3 µg/mL for ADA trough concentrations were identified. The cutoff point related to DR was lower than that observed in others studies15-18 which propose a therapeutic window of 5-10 µg/mL for IFX and 8-12 µg/mL for ADA to achieve MH. However, the trough concentrations in these studies were not obtained at steady state, which could explain the higher therapeutic range proposed. In this respect, cutoff values in the literature are very heterogeneous and other studies proposed values of 3.4 µg/mL19 for IFX and 4.9 µg/mL for ADA20. In this study, a huge percentage of patients with previous IBD surgery were included, and no differences in DR were observed, despite having lower anti-TNF concentrations. Anti-TNFs have shown high efficacy in preventing postoperative endoscopic recurrence of CD21, Sorrentino et al.22 suggested that lower doses of IFX, despite obtaining lower concentrations, maintained clinical and endoscopic remission in this patient profile. Although our study identified a specific cutoff point associated with DR, the range of concentrations associated with deep remission was wide, ranging from 0 to 14.4 µg/mL and from 2.36 to 15.6 µg/mL in IFX and ADA, respectively, so the target concentration should be individualized. Infliximab level up to 3.1 µg/mL resulted in an increased DR rate, and higher drug levels were not associated with a much greater increase in DR. In the case of ADA, major levels (>12 µg/mL) lead to improve DR proportion, so some patients may require even higher concentrations to achieve better results. In this regard, Plevris et al.23, explain that higher concentrations of ADA are necessary for fistula closure. However, in our study, only 6 patients achieved these concentrations, so no conclusions can be drawn.

Similarly, to the available evidence, CRP < 5 mg/L24 and FC < 100 µg/g25 were identified as independent factors of DR. In addition, high BMI, absence of previous IBD-related surgery, CRP < 5 mg/L, FC < 100 µg/g in IFX, and coadministration of IMMs, low BMI, absence of previous IBD-related surgery, CRP < 5 mg/L, and FC < 100 µg/g in ADA positively influence to achieve higher trough concentrations. The influence of BMI was in agreement with other studies, in a pharmacokinetic study26 of 24 patients receiving IFX, higher BMI correlated with higher post-infusion concentrations; although this did not correlate with a higher response rate, suggesting that circulating concentrations of the drug did not correlate with tissue concentrations. In the case of ADA, in a prospective pharmacokinetic study of Sanchez et al., BMI showed influence by increasing the apparent clearance of ADA, and was a significant variable in the pharmacokinetics of ADA27.

In contrast to other studies28, concomitant immunomodulators did not correlate with IFX concentrations in our study. However, in a multicenter retrospective study29 withdrawal of immunomodulators after at least 6 months of concomitant treatment with IFX did not reduce minimal IFX concentrations in patients with CD. In this regard, the patients included in our study had a median IFX treatment duration of 10.9 months, and could influence the lack of impact of immunomodulator coadministration on IFX concentrations. Coadministration of IMMs was associated with higher ADA concentrations in our series of patients, similar to a Diamond trial30, which evaluated the efficacy of ADA with and without azathioprine in CD patients, although not statistically significant (p 0.084), coadministration of IMMs tended to a higher minimum ADA concentration.

Immunogenicity was numerically higher in IFX than in ADA, as only 7 (6%) in IFX and 5 patients (4%) in ADA developed antibodies to the drug. However, the percentage of immunogenicity observed in the literature is very heterogeneous and can fluctuate depending on the immunoassay method used, varying widely between 0% and 79% in IFX and 0% and 87% in ADA, according to a recent review31. Positive ATI and ATA have been associated with a loss of clinical response32,33. In our study, ATA was negatively related to DR in univariate analysis, although it was not significant in multivariate analysis. Furthermore, only 1 of the 7 IBD patients with ATI, and no patients with ATA were in DR.

The use of a drug-sensitive assay that only detects concentrations of free TNF-α inhibitors and free anti-drug antibodies, with an inability to measure anti-drug antibodies in the presence of TNF-α inhibitors, could influence the low prevalence of patients with positive immunogenicity. Drug-tolerant assays from other manufacturers11 use an acidic buffer that dissociates immunocomplexes between anti-drug antibodies and ADA or IFX, allowing measurement of anti-drug antibodies that are part of these immunocomplexes. TDM using these kits could improve the sensitivity of immunogenicity detection, improving the analysis of the influence of drug antibodies on DR. However, in a post hoc analysis of the TAXIT34 trial that analyzed the clinical relevance of ATIs measured by a drug tolerance assay, it concluded that only high concentration ATIs, which are usually also detected in a drug sensitivity assay, persist over time and are clinically relevant and drug switching may be more cost-effective.

The main strength of our study was the systematic evaluation of DR considered as the new therapeutic target, using biomarkers and clinical and endoscopic scales. Unlike other studies, which evaluated the impact of IFX or ADA TDM using only biomarkers, or clinical and endoscopic assessment without standardization. Importantly, we were able to identify an optimal cutoff value for minimal IFX and ADA concentrations.

Our study has several weaknesses that should be mentioned. It is a single-center, real-world retrospective study, so confounders cannot be definitively excluded. No centralized reading of endoscopies was performed and interobserver variability cannot be excluded. Sampling was performed at different times in the course of the disease and a high percentage of patients with an intensification regimen were included. This may lead to an underestimation of the cut-off points, since concentrations that, due to a poor response, implied an intensification of treatment were not evaluated. In addition, clinical outcomes, such as hospitalization or colectomy, are not analyzed. The mean albumin level, a biomarker of anti-TNF clearance and enteropathy, was normal (4 g/L) in our cohort, and was not shown to be a factor associated with plasma concentrations or deep remission.

In conclusion, our findings highlight that IFX and ADA trough concentrations, CRP < 5 mg/L and FC < 100 µg/g are associated significantly with DR during maintenance therapy. Noteworthy, an optimal cut-off of 3.1 for IFX and 6.3 µg/mL for ADA were identified for predicting this outcome.

No funding.

No conflict of interest.