B. Verstockt1, J. Stylli2, P. Rahimian2, A. Rajulapati2, K. Ehring2, D. Thomas3, J. Sabino1, M. Ferrante1, S. Singh2, S. Vermeire1
1IBD Leuven, 1University Hospitals Leuven, Department of Gastroenterology and Hepatology, Leuven, Belgium, 2Progenity Inc., Progenity Inc., San Diego California, USA, 3KU Leuven, Laboratory for Therapeutic and Diagnostic Antibodies- Department of Pharmaceutical and Pharmacological Sciences, Leuven, Belgium
Background
Various studies report robust associations between serum adalimumab (ADM) exposure and therapeutic outcome, which has led to the implementation of reactive therapeutic drug monitoring. As very little is known how serum levels relate to tissue and faecal ADM levels and to tumour necrosis factor (TNF) levels, we measured serum, tissue and faecal drug and TNF, and questioned how all this is linked to the therapeutic efficacy of ADM.
Methods
We prospectively recruited paired serum, mucosal biopsies and stools from 15 patients with ulcerative colitis (UC, n = 8) and Crohn’s disease (CD, n = 7) initiating ADM. All received ADM standard dosage. Samples were collected at baseline and week 8 (UC) or month 6 (CD). Endoscopic remission was defined as a Mayo endoscopic sub-score 0 (UC) or complete absence of ulcerations (CD). ADM levels were measured using a homogeneous immunoassay in which the ADM-TNFα complex is formed and captured. TNFα and IL-6 were quantified using an in-house developed immunoassay. Total tissue protein content was quantified in order to normalise the data.
Results
Pre-treatment tissue TNF and IL6 levels correlated significantly (r = 0.8, p < 0.001) and were lower in ADM remitters (n = 8) in contrast to non-remitters (4.5 vs. 8.6 pg/mg, p = 0.03; 7.1 vs. 122.5 pg/mg, p = 0.05). Pre-treatment faecal TNF was numerically lower in remitters (0.4 vs. 1.0 pg/mg, p = 0.30), but did not correlate with tissue/serum TNF (p = 0.84, p = 0.58). Although ADM did not significantly impact post-exposure serum or faecal TNF and IL6, it did reduce tissue TNF and IL6 (p = 0.02, p = 0.003), both in remitters (median −38.3%, p = 0.04; median −86.0%, p = 0.04) and non-remitters (median −59.5%, p = 0.16; median −84.5%, p = 0.03). Although TNF and IL6 changes did not correlate with ADM exposure or change in calprotectin, the decrease in faecal calprotectin significantly correlated with tissue, but not serum, ADM levels (r = 0.74, p = 0.01). ADM exposure was significantly higher in remitters as compared with non-remitters, both in serum (14.6 vs. 5.9 mg/ml, p = 0.009) and in tissue (53.1 vs. 24.1 ng/mg, p = 0.04). However, serum and tissue ADM levels correlated only moderately (r = 0.5, p = 0.1). Finally, tissue TNF/tissue ADM ratios were significantly higher in non-remitters, both considering pre (p = 0.007) and post TNF (p = 0.03). Faecal loss of ADM could be documented in a single patient only.
Conclusion
Both in tissue and serum ADM exposure was significantly higher in remitters, though tissue and serum ADM exposure did not perfectly correlate. A higher baseline tissue TNF burden was associated with non-response to ADM, and may help in the differentiation to select the appropriate biological in a given patient: no anti-TNF or intensified dosing in patients with a high baseline tissue TNF burden.
