Anti-TNFα therapy with infliximab is effective for Crohn’s disease. Infliximab neutralizes the biological activities of TNFα, a cytokine involved in host-defence against certain infections.
Anti-TNFα therapy with infliximab is effective for Crohn’s disease. Infliximab neutralizes the biological activities of TNFα, a cytokine involved in host-defence against certain infections.
To evaluate the effects of infliximab on the gut and peripheral immune system functions.
Biopsies and blood samples from three clinical trials of infliximab in Crohn’s disease were analysed. Pharmacokinetics, changes in leucocyte counts and T cell subsets, T cell function, and cytokine profiles of lamina propria mononuclear cells (LPMC) and peripheral blood mononuclear cells (PBMC) were analysed.
Infliximab has a serum half-life of 9.5 days and is still detectable in serum 8 weeks after infusion. Leucocyte counts showed consistent changes from baseline toward normal values after therapy. Monocytes and lymphocytes were modestly increased, while neutrophils were decreased 4 weeks after treatment. Lymphocyte subsets and T cell proliferative responses were not altered after therapy. The proportion of PBMCs capable of producing IFNγ and TNFα did not change, while Th1 cytokine production by stimulated LPMC was decreased after infliximab therapy.
The clinical efficacy of infliximab is based on local anti-inflammatory and immunomodulatory effects in the bowel mucosa, without generalized suppression of systemic immune functions in Crohn’s disease patients.
Infliximab (chimeric anti-human TNFα, Remicade) induces profound and durable responses in a large number of patients with moderate to severe and fistulizing Crohn’s disease.1, 2 The long duration of response suggests that infliximab’s mechanism of action may be more complex than simply neutralizing soluble TNFα.
Local effects of infliximab on the inflamed bowel mucosa have been observed. Thus, the number of TNFα expressing cells is reduced 4 weeks after treatment with infliximab in Crohn’s colitis. CD4+ and CD8+ cells (the major T cell subtypes), and CD68+ monocytes and macrophages, all important producers of TNFα, are reduced by > 50% in the colonic lamina propria of infliximab-treated patients but not in placebo-treated patients.3 The number of IFNγ and TNFα-producing mononuclear cells isolated from the lamina propria in patients with Crohn’s disease is decreased after treatment with infliximab.4 Infliximab reduces the expression of the adhesion molecules ICAM-1 and LFA-1.3 Together with histological repair, remarkable endoscopic healing has been observed, and a correlation between endoscopic healing and reduction of disease activity has been demonstrated after infliximab therapy.5 The profound clinical, biological, histological, and endoscopic effects of infliximab in patients with Crohn’s disease may be explained by the rapid and effective neutralization of TNFα by infliximab, although lysis of activated macrophages which display cell surface transmembrane TNFα may also be important.
The impact of infliximab on systemic immunity is not clear. While TNFα can influence the inflammatory process and modulate the immune response to specific pathogens, the primary defect reported in mice deficient in TNFα is altered cell organization in secondary lymphoid tissues.6 Although mild infections such as upper respiratory tract infections are increased in infliximab-treated patients, opportunistic infections were not observed.7 Therefore, it is possible that infliximab exerts significant effects on the pathologic mucosal immune response in Crohn’s disease, while systemic immune responses remain intact.
This study reviews data collected during several clinical studies of infliximab in Crohn’s disease and aims at evaluating the effect of infliximab on the local (bowel mucosa) and systemic immune functions in treated patients.
This paper describes findings from previous clinical trials of infliximab in Crohn’s disease.
The T11 study included 20 patients with active Crohn’s disease. This was an open-label, dose-ranging study which enrolled five patients per dose group receiving a single infusion of 1, 5, 10 or 20 mg/kg of infliximab. Blood samples were obtained before infliximab treatment and at 1, 2, 4, 24 and 72 h and 1, 2, 4, 8 and 12 weeks following infliximab treatment.
The T16 study was a double-blind, placebo-controlled trial that included 108 patients with moderate to severe, active and refractory Crohn’s disease. Patients were treated with a single infliximab infusion of 5 mg/kg (n=27), 10 mg/kg (n=28), 20 mg/kg (n=28) or placebo (n=25). Serum samples were obtained prior to and after infusion of infliximab at 1, 2 and 4 h and at 2, 4, 8 and 12 weeks.
The T20 study was a double-blind, placebo-controlled trial performed in 94 patients with fistulizing Crohn’s disease not responding to conventional therapy. Three intravenous infusions of 5 mg/kg (n=31) or 10 mg/kg (n=32) infliximab or placebo (n=31) were given at weeks 0, 2 and 6. Blood samples were collected prior to and 1 h after each infusion, and at weeks 10, 14 and 18 after the first infliximab treatment.
Serum concentrations of infliximab were measured in T16 and T20. TNFα was measured in serum samples collected in T11. Both IL-6 and C-reactive protein (CRP) were evaluated in serum samples from T16. Peripheral white blood cell counts (WBC) were performed in both T16 and T20. The proliferative response of peripheral blood mononuclear cells (PBMC) was analysed in a subset of patients from T16 and T20. Cytokine production by PBMC was evaluated in a subset of patients in T16, and in lamina propria mononuclear cells (LPMC) isolated from a subset of patients in T11 and T16. Pre- and post-treatment samples were compared in individual patients, and diseased tissue was compared with normal tissue from uninvolved areas in the same patient.
Serum concentrations of infliximab were determined using a monoclonal antibody-based enzyme immunoassay (detection limit 0.1 μg/mL). Serum concentrations of TNFα were measured using a commercially available immunoassay kit (detection limit < 3 pg/mL; Biosource, Belgium).
CRP serum concentrations were measured by rate nephelometry (detection limit 0.1 mg/dL). Interleukin (IL)-6 concentrations were determined by enzyme immunoassay (detection limit 0.1 pg/mL; R&D Systems).
Analyses of WBC counts were incorporated into the haematology evaluations that were routinely conducted before and after infliximab treatment in each clinical trial. Lymphocyte subset measurements were performed in patients before and 6 weeks after infliximab infusion. Fifty microlitres of heparinized blood were incubated at 4 °C with an appropriate dilution of a fluorescein isothiocyanante (FITC)- or phycoerythrin (PE)-conjugated monoclonal antibody (Becton-Dickinson, CA, USA) specific for lymphocyte surface markers. After haemolysis of erythrocytes (twice-repeated addition of NH4Cl, pH=7.3), the cells were washed twice in PBS and re-suspended in 1 mL of 1% paraformaldehyde. Combinations of monoclonal antibodies were selected in order to determine the following lymphocyte subsets:
1 T lymphocytes: CD3+ and CD5+ total T cells, α/β+ and γ/δ+ T cells, CD4+ helper/inducer and CD8+ suppressor/cytotoxic T cells, CD8+/CD28+ (CD28 binds to B7 on APCs), CD3+/CD56+ and CD57+ (CD56 and CD57 are markers of NK cell activity on non-T cells, and for long-term memory on T cells), CD3+/CD25+ (CD25 is the IL-2 receptor α-chain) T cells and total HLA-DR+, CD3+/HLA-DR+, CD4+/HLA-DR+, CD8+/HLA-DR+.
2 B lymphocytes: CD19+ and CD19+/CD5+ B cells.
3 Non-T non-B lymphocytes: total, CD56+/CD16+, CD57+, CD8+.
Monocytes were counted by immunofluorescence with PE labelled anti-CD14. Two-colour immunofluorescence analysis was carried out on a Becton–Dickinson fluorescence-activated cell sorter (FACS). Lymphocytes were gated on the basis of forward and 90° light scatter, and 2000–5000 cells were counted in each analysis. Cells were scored positive or negative according to their fluorescence, with reference to that of control samples, incubated with FITC- and PE-conjugated mouse IgG. The number of positive cells was expressed as a percentage of the total lymphocyte count.
PBMC were isolated from 30 mL of heparinized blood drawn prior to and 4 weeks after infusion of infliximab on Ficoll-Hypaque (density 1.077) gradients (Pharmacia, Uppsala, Sweden). After three washings in Hanks’ balanced salt solution, the cells were re-suspended in RPMI 1640 culture medium (Gibco, Paisley, UK) supplemented with 2 m M L-glutamine, penicillin (100 U/mL), streptomycin (100 μg/mL), and 5% normal human serum. To investigate the proliferative response of PBMC to antigenic and mitogenic stimuli, the cells (1 × 105) were cultured in 96-well round-bottomed plates (Costar Europe, Badhoevedorp, Netherlands) with a variety of soluble antigens and mitogens. The following antigens were used: Candida albicans (10 μg/mL, Haarlem Allergenen Laboratorium, Netherlands), varidase (100 U/mL, Lederle, Cyanamid Benelux, Brussels, Belgium), tuberculin (20 U, Statens Serum Institute, Copenhagen, Denmark) and heat shock protein-65 (HSP-65, 5 and 10 μg, a gift from J. Van Embden, Bilthoven, Netherlands). The following mitogens were used: pokeweed mitogen (PWM, 0.5 μg/mL, Sigma, St Louis, MO, USA), anti-CD3 (5 μg/mL, polyclonal stimulator UCTH-1, a gift from P. Beverley, Imperial College, London, UK), IL-2 (50 U/mL, Boehringer Mannheim), anti-CD3 plus IL-2 (50 U/mL), anti-CD3 plus phorbol myristate acetate (PMA, 25 ng/mL, Calbiochem Novabiochem), and PMA (25 ng/mL) plus ionomycin (Ca-ionophore 1 μg/mL, Calbiochem Novabiochem).
Cultures were incubated in quadruplicate at 37 °C in a 5% CO2 atmosphere for 6 days, and 8 h after a 1 μCi (3H)thymidine pulse (2 Ci/mmol; Amersham, Buckinghamshire, UK) cells were harvested and processed for the determination of (3H)thymidine incorporation. Incorporation by cells cultured in medium alone was substracted from the values for antigen and mitogen stimulated incorporation.
In vitro release of TNFα, IFNγ and IL-10 by PBMC stimulated with concanavalin A (Sigma, St Louis, MO) and PMA (Sigma) were studied before and at 1, 2, 3 and 4 months after infliximab treatment, as previously described.4
Mononuclear cells from bowel lamina propria were isolated before and 1, 2, 3 and 4 months after infliximab treatment from biopsies of inflamed areas, as well as from uninvolved (visually normal) areas of colon. The number of cells stimulated to secrete these test cytokines was quantified by reversed enzyme linked immunospot assay (relispot assay)4 and compared to matched, unstimulated cells. TNFα, IFNγ and IL-10 release in vitro by unstimulated (control) and anti-CD2 and anti-CD28 stimulated LPMC was quantified.
A dose-dependent maximum serum concentration (Cmax) of infliximab was observed after a single infusion of 5 mg/kg, 10 mg/kg or 20 mg/kg infliximab (Figure 1). At the 5 mg/kg recommended dose the median Cmax was 118 μg/mL, and the terminal half-life of infliximab was 9.5 days. Clearance of infliximab from the circulation was 9.8 mL/h for the 5 mg/kg dose and was similar for the other doses studied. At week 12, infliximab was no longer detectable (median concentration < 0.1 μg/mL) in the 5 mg/kg dose group.
Further analysis showed that there was no difference in the serum levels of infliximab between clinical responders (defined as a ≥ 70 point reduction in the Crohn’s disease activity index 4 weeks after infliximab treatment) and non-responders. Median serum levels were 99.5 μg/mL and 118 μg/mL at 2 h post-infusion of 5 mg/kg infliximab, 28.6 μg/mL and 29.7 μg/mL at 2 weeks, and 10.9 μg/mL and 10.6 μg/mL at 4 weeks in responders and non-responders, respectively. Therefore, the difference in response to treatment cannot be explained by a difference in pharmacokinetics.
Re-administration of infliximab at weeks 2 and 6 (as in the T20 study of fistulizing patients) restored the initially achieved serum infliximab concentrations. After the first, second and third infusions of infliximab, the median serum concentrations were 158 μg/mL, 195 μg/mL and 178 μg/mL, respectively, for the 5 mg/kg dose group (Figure 2).
Assays of TNFα in serum revealed low levels (5–25 pg/mL) of TNFα at baseline, followed by a rise in TNFα from 4 to 72 h. Peak concentrations of TNFα were detected at 72 h, to 2 weeks, which then decreased to baseline by week 12 (Figure 3). This immune reactivity in the TNFα assay detected after infusion of infliximab was shown to be TNFα-complexed with infliximab.8 This was confirmed by testing the 72 h serum samples in a WEHI cell cytotoxicity assay and demonstrating that there was no bioreactive TNFα in these samples (data not shown).
The acute phase inflammatory response is an important TNFα-inducible element of the immune system and can be monitored by measuring CRP and IL-6 in serum samples. A single infusion of 5 mg/kg infliximab results in normalization of CRP levels by 2 weeks, which is maintained through 12 weeks (Figure 4a), while a gradual increase was observed for patients treated with placebo. IL-6 serum levels were also normalized at 2 weeks after infusion of 5 mg/kg, then gradually increased in the 5 mg/kg group, but remained within the normal range (1.6 pg/mL) in the higher dose groups (Figure 4b). No changes in IL-6 concentrations were observed in the placebo-treated group. Similar results were obtained in samples from T20 for both CRP and IL-6 serum levels (data not shown).
The changes in differential cell counts for the 18-week period in study T20 are shown in Figure 5. Eosinophils and basophils comprised minor fractions of the total cell counts and did not show systematic changes over time. Baseline counts of monocytes, lymphocytes and neutrophils were 449, 1.452 and 6.424/mm3, respectively (5 mg/kg dose group of T20). Increases in the median percentage of both monocytes and lymphocytes occurred in the infliximab treatment groups. In the 5 mg/kg infliximab group, the median lymphocyte percentage increased from 16.5% at baseline to 26.5% at 10 weeks (to 2.332/mm3) while monocytes increased from 5.1% to 7.0% (to 616/mm3) in the same time period. A corresponding decrease in the median percentage of neutrophils accompanied the increased lymphocyte fraction (from 6.424 to 5.430/mm3). Both lymphocyte and neutrophil fractions returned toward the baseline profile at 14 and 18 weeks (8 and 12 weeks after the last infliximab infusion). Similar results were obtained from the T16 study (data not shown). The changes in differential cell counts were not due to altered total WBC counts (from 8.800/mm3 at baseline to 8.100/mm3 at week 18 for the 5 mg/kg dose group of T20; Figure 6). A modest decrease (T16) or no change (T20) was observed following single or multiple infliximab infusions, respectively.
Peripheral blood cells were incubated with FITC or PE-conjugated antibodies specific for lymphocyte cell surface markers and analysed by FACS. As shown in Table 1, there were no marked trends observed in a comparison of the baseline and 6-week values in the infliximab-treated or placebo-treated patients, or by a comparison of the infliximab-treated and placebo groups. In addition, no changes in expression of activation markers (e.g. HLA-DR, CD57) were observed.
T cells were studied from patients treated with single (T16) or multiple (T20) infliximab infusions. The results of these proliferation assays are summarized in Figure 7. Proliferation in response to specific antigens was not altered by infliximab treatment. Non-specific mitogens also showed no substantial changes from baseline, with the exception of ionomycin plus PMA. Response to ionomycin plus PMA increased proliferative responses approximately 2.3-fold in the infliximab group compared to a slight decrease in the placebo group.
IL-12 induces IFNγ, a marker of Th1 response and activator of macrophages. TNFα stimulates Th1 development and inhibits Th2 cell proliferation, while IL-10 blocks the activation of Th1 cells, and is considered a natural inhibitor of inflammation. Therefore, the capacity of in vitro stimulated lymphocytes to produce IFNγ, TNFα and IL-10 provides information on the overall immune status of individuals. As shown in Figure 8, all three cytokines are increased in the stimulated PBMC compared to unstimulated control cell cultures at all time points. The in vitro production of IFNγ and TNFα by stimulated PBMC does not reveal any systematic trend after treatment with infliximab. This result suggests that the circulating population of lymphocytes capable of producing these cytokines is not affected by treatment with infliximab. With IL-10, there was a trend towards lower levels of production, particularly at later time points.
Biopsies were taken from inflamed areas, as well as from uninvolved (visually normal) areas of the colon. Isolated cells were incubated in the absence (control) or presence of anti-CD2 and anti-CD28 antibodies. The results, expressed as cytokine positive cells per 10 000 total cells, are shown in Figure 9. The number of cells from inflamed areas that could be induced to secrete Th1 cytokines (IFNγ and TNFα) decreases between 1 and 2 months and remained low through 4 months. Interestingly, the biopsies from uninflamed areas, although endoscopically normal, also had substantial numbers of cells that could be induced to produce IFNγ and TNFα and which also declined between 1 and 2 months after infliximab treatment. IL-10, which was not induced after stimulation of the LPMC, showed a trend toward lower number of cells producing this cytokine over the entire 4-month period of the study in both inflamed and uninflamed biopsies, although IL-10 producing cells in all biopsies were rare.
The pharmacokinetics of infliximab were found to be predictable and consistent over all three clinical trials in Crohn’s disease. The terminal half-life of infliximab of 9.5 days is comparable with that of other IgG molecules.9 Because of its long half-life and slow clearance rate (9.8 mL/h), infliximab remains detectable in serum for at least 8 weeks after a single infusion of 5 mg/kg. The prolonged bioavailability of infliximab, together with its high affinity (Ka=1010 M−1)10 for TNFα results in effective and prolonged blockade of TNFα in patients. The median time to relapse in moderate to severe and fistulizing Crohn’s disease patients is 2–3 months. Recently, Rutgeerts et al.7 have shown that patients with moderate to severe Crohn’s disease with an initial response to infliximab retain response and remission when re-treated every 8 weeks.
Measurements of free TNFα are hampered by the presence of TNFα:infliximab complexes in serum. However, this confirms the in vivo binding of TNFα by infliximab, and the circulation of these complexes for up to 12 weeks after a single infusion.8 The experience in Crohn’s disease patients provides clinical evidence that complexed TNFα is no longer bioactive. The immediate effect of TNFα neutralization by infliximab is evidenced by a rapid decrease of CRP and IL-6 levels. These inflammatory markers were normalized by 2 weeks post-infusion, but other (unpublished) observations have demonstrated that these markers normalize within 48 h after administration of infliximab. Rapid onset of action is an important therapeutic advantage of infliximab compared with immunosuppressive therapies such as 6-mercaptopurine, for which the onset of response is often delayed and requires more than 4 months in 19% of patients.11 Infliximab blocks IL-6, and thereby inhibits the production of the acute phase reactant CRP by the liver. When the signs and symptoms of disease return, usually after a 2–3-month period, both IL-6 and CRP increase. CRP concentrations remained below the normal range (0.8 mg/dL) in patients retreated every 8 weeks during a 36-week period, but not in placebo-retreated patients.7
The percentage of circulating lymphocytes and monocytes is increased between weeks 2 and 14 after infliximab infusions in T20, whereas the fraction of neutrophils is decreased. No significant changes were seen in the percentages of T cell subsets at 6 weeks after infliximab treatment. Recently, Maurice et al.12 have described increased numbers of circulating CD4+ and CD8+ T cells in the blood of rheumatoid arthritis patients 3 days after treatment with infliximab. They suggest that homing of Th1 cells (secreting proinflammatory cytokines) into the inflamed synovium of rheumatoid arthritis patients is blocked by infliximab and that, as a result, these cells accumulate temporarily in the peripheral circulation.12 Our data cannot confirm these results, as we focused on long-term rather than short-term effects.
Immunohistochemical studies have shown a significant reduction in the number of activated T cells in the lamina propria of the ileum and colon of Crohn’s disease patients treated with infliximab.3 The percentage of cells with positive stainings for ICAM-1 and LFA-1 were also significantly decreased.3 Decreased numbers of inflammatory Th1 cells and TNFα positive monocytes and macrophages after infliximab therapy may be explained by increased in situ cell lysis, together with decreased cell influx due to down-regulation of adhesion molecules within the inflamed tissue. This study suggests that the homing of Th1 cytokine secreting T cells may also be inhibited by infliximab in Crohn’s disease.
Our results demonstrated modest declines in the total WBC count following treatment with infliximab. However, the median WBC counts at all time points were within the normal range for adults (4500–11 000/mm3). Mild leukopenia and more severe pancytopenia, while transient and reversible, have been reported with 6-MP11 and azathioprine,13 respectively. Similar changes have not been observed with infliximab. No prolonged changes have been observed in circulating lymphocyte subsets. The anti-inflammatory action of infliximab has local effects at the tissue level, resulting in architectural repair and tissue healing,5 as well as systemic effects on the signs and symptoms of the disease, as shown by decreases in CDAI and increases in IBDQ scores. The immune modulatory action of infliximab is mainly apparent at the tissue level, while the systemic immune function remains largely unchanged.
Proliferation assays of T cells performed 4 weeks after an infusion of infliximab demonstrate that these cells are immune-competent in response to antigenic and mitogenic challenges. The antigen-specific and mitogen responses indicate that infliximab treatment does not induce a generalized suppression of cellular immune functions, and that with certain stimuli, such as ionomycin plus PMA, enhanced proliferative responses can be seen. This is in agreement with the earlier findings of Cope et al.14 which demonstrated improved proliferative responses of lymphocytes from rheumatoid arthritis patients after treatment with infliximab.
The in vitro production of the proinflammatory cytokines TNFα and IFNγ and the anti-inflammatory cytokine IL-10 by PBMC was studied after treatment with infliximab. Stimulation of PBMC with concanavalin A plus PMA resulted in increased INFγ and TNFα secretion, with little change observed over the 0–4 month post-infusion period. For IL-10, a trend towards lower production at later time points was evident. These results suggest that circulating Th1 lymphocytes are fully functional following infliximab treatment. We did not analyse the in vitro production of typical Th2 cytokines such as IL-4, IL-5 or IL-13 by PBMC after infliximab therapy. IL-10 is produced by both Th1 and Th2 cells and also by macrophages. The source of IL-10 measured in our in vitro assays is uncertain, and a possible effect of infliximab on Th2 cytokine production could therefore not be assessed.
In Crohn’s disease patients responding to infliximab therapy, Plevy et al.4 have shown that the numbers of LPMC-producing TNFα and IFNγ following CD2/CD28 activation are decreased over an 8-week period. In addition, this decrease in Th1 LPMC number was correlated with a decrease in CDAI (response) and endoscopic score. We confirm and further extend these findings. After a single infusion of infliximab the fraction of Th1 positive LPMC is decreased through 4 months after infusion in both inflamed and uninflamed areas. These results confirm that infliximab therapy may down-regulate proinflammatory cytokine responses in inflamed and uninflamed Crohn’s mucosa.4 Interestingly, Plevy et al.4 have also shown that CD2/CD28-activated IFNγ and TNFα production from LPMC is down-regulated following infliximab therapy, while the production of IFNγ by PBMC is increased in clinically responsive patients. Therefore, and in agreement with findings in rheumatoid arthritis patients,12 one effect of infliximab may be to prevent homing of Th1 cells from the periphery to the lamina propria.
In conclusion, the data presented from these clinical studies demonstrate that infliximab has profound local and systemic anti-inflammatory effects. The immune modulatory effects of infliximab are driven by the action of TNFα in the inflamed mucosa. Down-regulation of adhesion molecules within the lamina propria may inhibit homing of Th1 cells, resulting in an accumulation of lymphocytes in the circulation. Our results show no evidence of a generalized suppression of the immune system after infliximab therapy.