Changes in T Lymphocytes and Cytokines After Anti-TNF Treatment in Pediatric Inflammatory Bowel Disease: Association with Response to Pharmacologic Therapy

Abstract

Failure of anti-TNF therapy is a real concern in children with inflammatory bowel disease (IBD) owing to the limited therapeutic arsenal. Anti-TNF drugs modulate the immune response, a key driver of chronic inflammation in IBD. Accordingly, we analyzed changes in the frequency of T-lymphocyte and cytokine levels after 6 weeks of treatment to identify potential biomarkers of response to anti-TNF drugs. We recruited 77 patients under 18 years of age diagnosed with IBD and treated with an anti-TNF drug. Using flow cytometry and multiplex ELISA, we analyzed 31 T-lymphocyte populations and four cytokines. We identified changes in 10 populations of T lymphocytes after 6 weeks of treatment. Naïve Tregs were associated with a primary response to anti-TNF drugs, while activated Tregs were associated with long-term response. Serum INF-γ levels were decreased after anti-TNF treatment in children with Crohn's disease (CD), but not in those with ulcerative colitis (UC). The memory CD8+ Type 2 Cytotoxic T (Tc2) subset increased in non-responders with CD and the CD4+ memory Th17 cells increased in non-responders with UC. These findings could help us to understand the cellular regulation of anti-TNF therapy, to identify children at a higher risk of treatment failure, and, potentially, to develop more personalized therapeutic strategies.

Keywords: T-lymphocytes; Treg; adalimumab; cytokines; inflammatory bowel disease; infliximab; pediatric; response biomarkers.

Figure 1

Flow chart showing the selection of study patients. The patients included in each study were divided into week 0 and week 6 and further classified into primary responders and primary non-responders. No response data were obtained for 1 patient included in the cytokine study. R, responders; NR, non-responders; UR, unclassified response.

Figure 2

Frequency of T-cell populations before and after 6 weeks of treatment with an anti-TNF drug: (a) total lymphocytes (gated on leukocytes); (b) naïve CD4+ (gated on CD4+ T-cells); (c) CM CD4+ (gated on CD4+ T-cells); (d) EM CD4+ (gated on CD4+ T-cells); (e) Th0 (gated on CXCR3- CCR4- CD4+ T-cells); (f) Th1 (gated on CXCR3+ CCR4- CD4+ T-cells); (g) Th9 (gated on CD4+ T-cells); (h) Th17 (gated on CXCR3- CCR4+ CD4+ T-cells); (i) Th1/Th17 (gated on CXCR3+ CCR4- CD4+ T-cells); (j) RTE Treg (gated on Treg cells). * p value < 0.05; ** p value < 0.01; *** p value < 0.001.

Figure 3

Frequency of T-cell populations in responders and non-responders after 6 weeks of treatment with an anti-TNF drug: (a) Naïve Tregs (gated on Treg cells); (b) EM Tregs (gated on Treg cells). * p value < 0.05.

Figure 4

A low frequency of Act Treg cells (gated on Treg cells) after 6 weeks of anti-TNF treatment is associated with long-term failure of anti-TNF drugs. (a) Comparison of the frequency of Act Treg in patients whose anti-TNF therapy failed (NR) or did not fail (R) during the follow-up; (b) Kaplan–Meier curve of patients with high (>8.030%), medium (3.205% to 8.030%), and low (<3.205%) frequency of Act Treg cells. * p value < 0.05.

Figure 5

Levels of cytokines prior (T0) and after 6 weeks of treatment (T6) with anti-TNF therapy: (a) IFN-γ; (b) IL-10; (c) IL-17A; (d) IL-4. *** p value < 0.001.

Figure 6

Levels of cytokines stratified by time of administration of the anti-TNF agent, namely, week 0 (T0) or week 6 (T6), and type of disease, namely, CD or UC. (a) IFN-γ; (b) IL-10; (c) IL-17A; (d) IL-4. * p value < 0.05; ** p value < 0.01; *** p value < 0.001.

Figure 7

Use of unsupervised algorithms to detect non-conventional cellular subsets. (a) Box plot histogram showing the abundance of metacluster_18 in responders (R) and non-responders (NR) among individuals with CD at T0 and (b) T6 obtained through opt-SNE and FlowSOM analysis. (c) Arbitrary unit of the medians of fluorescence (MFI) of cellular markers in the metacluster_18. (d) Box plot histogram showing the abundance of metacluster_40 in responders and non-responders among individuals with UC at T0 and (e) T6 obtained through opt-SNE and FlowSOM analysis. (f) Arbitrary unit of the MFI of the cellular markers of metacluster_40. Intermediate dotted line, mean; intermediate unbroken line, median; the upper and lower boxes represent the first and third quartiles, respectively. P corresponded to the adjusted p value.

Figure 7

Use of unsupervised algorithms to detect non-conventional cellular subsets. (a) Box plot histogram showing the abundance of metacluster_18 in responders (R) and non-responders (NR) among individuals with CD at T0 and (b) T6 obtained through opt-SNE and FlowSOM analysis. (c) Arbitrary unit of the medians of fluorescence (MFI) of cellular markers in the metacluster_18. (d) Box plot histogram showing the abundance of metacluster_40 in responders and non-responders among individuals with UC at T0 and (e) T6 obtained through opt-SNE and FlowSOM analysis. (f) Arbitrary unit of the MFI of the cellular markers of metacluster_40. Intermediate dotted line, mean; intermediate unbroken line, median; the upper and lower boxes represent the first and third quartiles, respectively. P corresponded to the adjusted p value.