Tumour necrosis factor alpha down-regulation parallels inflammatoryregression in ulcerative colitis patients treated with infliximab
C. Hassan a,∗, E. Ierardi b, O. Burattini b, V. De Francesco b, A. Zullo a,G. Stoppino b, C. Panella b, S. Morini a
a Gastroenterology and Digestive Endoscopy Unit, “Nuovo Regina Margherita” Hospital, Rome, Italyb Section of Gastroenterology, Department of Medical Sciences, University of Foggia, Foggia, Italy
Received 4 February 2007; accepted 6 June 2007Available online 25 July 2007
bstract
Background. Treatment with the anti-tumour necrosis factor � monoclonal antibody infliximab has been shown to be effective in moderate-o-severe ulcerative colitis. However, its effect on the mucosal histopathological abnormalities of this disease is largely unknown. This studyimed to assess the immunohistological effect of infliximab in ulcerative colitis.
Methods. Nine patients with moderate-to-severe ulcerative colitis received infliximab (5 mg/kg) at weeks 0, 2 and 6, respectively. Coloniciopsies were collected before therapy and at week 10, when the Mayo score (including the endoscopic subscore) was also assessed. Severity ofnflammation was evaluated by histologic score and histomorphometry. Immunohistochemical staining with antibody against tumour necrosisactor � was performed on all biopsies and expressed as percentage of positive stromal cells/1000 counted (tumour necrosis factor � score).
Results. A profound down-regulation of mucosal tumour necrosis factor � occurred in all the six patients who achieved a clinical response,ut not in the three who did not respond. Median tumour necrosis factor � score dropped from 44.8 (range 35–58.3) to 12.8 (range 5.3–15.3)n the responders (p = 0.03), whilst it remained unchanged in the non-responders. Such effect was related with a dramatic regression of the
edian histologic score, which dropped from 2.7 (range 2–3) to 0.5 (range 0.0–1.5) in responder patients (p = 0.002). This was related tovirtual disappearance of neutrophils in responders (r = 0.72; p = 0.002; Spearman’s test), but not in those who did not improve. Tumour
ecrosis factor � score appeared to be correlated with the histologic, endoscopic and clinical activity.
Conclusions. A profound tumour necrosis factor � down-regulation appears to be strictly associated with a dramatic regression of the
nflammation in patients with moderate-to-severe ulcerative colitis treated with infliximab. Such immunohistochemical effect seems to beritical for a clinical and endoscopic response to therapy.
2007 Published by Elsevier Ltd on behalf of Editrice Gastroenterologica Italiana S.r.l.
lcerativ
pcd
eywords: Biologic therapy; Immunohistochemistry; Infliximab; TNF-�; U
. Introduction
Inflammatory bowel diseases have been shown to be the
esult of the expression of a genetically determined dysregu-ation of the host immune response toward luminal antigens,esulting in an immunological imbalance with an excess of
∗ Corresponding author at: Ospedale Nuovo Regina Margherita, Gastroen-erologia ed Endoscopia Digestiva, Via Morosini 30, 00153 Rome, Italy.el.: +39 06 58446541; fax: +39 06 58446533.
ro-inflammatory cytokines [1]. In detail, a type 1 helper Tell response has been claimed to play a major role in Crohn’sisease (CD), whilst a type 2 helper T cell response haseen implicated in the pathogenesis of ulcerative colitis (UC)2,3].
Tumour necrosis factor � (TNF-�) – a 17 kDa polypeptideroduced by macrophages, lymphocytes and natural killer
ells – has been mainly implicated in the pathogenesis ofype 1 helper T cell-mediated disorders, such as CD andheumatoid arthritis [4,5]. However, high levels of TNF-�ave also been shown in the blood, stool and intestinal tissues
f patients with UC, suggesting a possible role even in typehelper T cell-mediated diseases [6–8].Infliximab (Remicade®; Centocor) – the first and most
xtensively studied biological agent for inflammatory boweliseases – is a chimeric monoclonal antibody to TNF-� com-osed of a complement-fixing ‘human’ IgG1 constant region75%) and a murine-derived antigen-binding variable region25%), able to bind with high affinity to both soluble andembrane-bound TNF-� receptors [9]. Two large random-
zed, double-blind, placebo controlled trials have shown theigh efficacy and safety of infliximab in inducing and main-aining a clinical and endoscopic remission in patients with
oderate-to-severe active UC, similarly to the success previ-usly reported in moderate-to-severe luminal and fistulizingD [10–12]. In CD patients, a profound down-regulation ofNF-� in the affected mucosa has been associated with aramatic histological improvement 4 weeks after infliximabherapy, strengthening the importance of such cytokine in CDathogenesis [13]. Intriguingly, a similar decrease has beenlso shown in the synovial tissue of patients with rheumatoidrthritis treated with infliximab, suggesting that a profoundNF-� down-regulation is critical for the therapeutic success
n type 1 helper T cell disorders [14]. On the other hand, its still unknown whether in patients with UC – and generallyn type 2 helper T cell-related diseases – such a profoundown-regulation does occur. This evidence would be impor-ant not only to better understand the mechanism of action ofhe drug, but also to unveil the role played by TNF-� in UCathogenesis.
To address these issues, we performed a prospective studyn order to assess mucosal TNF-� levels before and afternfliximab treatment in consecutive patients with moderate-o-severe UC, and to associate such values with both thelinical response, and the endoscopic and histological activ-ty.
. Methods
.1. Patients
Consecutive patients with an indication to infliximabherapy because of steroid-dependant or steroid-refractory
oderate-to-severe UC (Mayo score ≥6) [15] were con-idered for enrolment. Exclusion criteria were a positiveuberculin skin test, chest X-ray alterations, a previous usef anti-TNF agents, or a diagnosis of indeterminate colitisr Crohn’s disease at the entry endoscopy. Infliximab wasdministered at weeks 0, 2 and 6, respectively at a standardose of 5 mg/kg. Mayo score was assessed at enrolment andt week 10, when endoscopy was also performed. Clinicalesponse at week 10 was defined as a decrease from baseline
n the total Mayo score of at least three points and at least0%, with an accompanying decrease in the subscore for rec-al bleeding of at least one point or an absolute subscore forectal bleeding of 0 or 1. Clinical remission was defined as
cmir
Disease 39 (2007) 811–817
total Mayo score of two points or lower, with no individ-al subscore exceeding one point [10]. For the purpose ofhe study, patients who achieved a clinical response or remis-ion were grouped altogether as responder patients. A writtennformed consent was obtained from all participants in thistudy.
.2. Endoscopy and histology
A flexible sigmoidoscopy, unless a colonoscopy was clin-cally indicated, was performed before the start of the therapynd at week 10. Mayo endoscopic subscore was used tossess endoscopic activity. Mucosal healing was defined asn absolute Mayo subscore for endoscopy of 0 or 1. Inach patient, at least four biopsy specimens per endoscopyere collected in the sigmoid tract. Biopsy specimens were
lways taken in the vicinity of lesions or ulcerations, ifresent. Two specimens were routinely processed and stainedith H&E; the remaining specimens were used for immuno-istochemistry. In detail, the H&E-stained specimens wereemiquantitavely scored according to the histological dis-ase activity. This score represents the mean value of theingle scores of the following histological characteristics:olymorphonuclear infiltration of the epithelium and laminaropria, crypt abscesses, loss of glandular parallelism, crypthortening and/or ramification, mucus epithelial depletion,nvolvement of muscularis mucosae and/or submucosa [16].ach histology score ranged from 0 (normal) to 3 (1 mild, 2oderate and 3 severe). In addition, the total number of neu-
rophils, lymphocytes, and plasma cells in the lamina propriaas counted in five high-power fields (original magnifica-
ion 400×), selected on the basis of high cellular density,nd it was expressed as number of cells/mm2 through a ded-cated software (Leica Q Win, Leica Microsystems Imagingolutions Ltd., Cambridge, UK).
.3. Immunohistochemistry
TNF-� expression was assessed by immunohistochem-stry, as we previously reported [17]. The cytokine wastained using a polyclonal rabbit antibody (PromoCell,eidelberg, Germany). The reaction was revealed using aeroxidase/anti-peroxidase (PAP) technique with goat anti-abbit immunoglobulins (Dako, Copenhagen, Denmark) and
complex of rabbit antibodies and horseradish peroxi-ase (Dako, Copenhagen, Denmark). A positive reactionas revealed by aminoethylcarbazole and counterstained by
queous haematoxylin. For each staining procedure we usedositive (Helicobacter pylori chronic gastritis with severectivity) and negative (normal stomach) controls. In themmunoperoxidase-stained biopsy specimens, TNF-� scoreas defined as the percentage of positive stained stromal
ells compared with the total number of lamina propriaononuclear cells, counted in five high-power fields (orig-
nal magnification 400×, at least 1000 total counted cells),andomly selected. Samples from an equal number of sub-
C. Hassan et al. / Digestive and Liver Disease 39 (2007) 811–817 813
Fig. 1. Histological changes in colonic biopsies before and after infliximab therapy (H&E; original magnification 400×). (A) Active UC characterized by adense inflammatory infiltrate of lymphocytes, plasmacells and polymorfonuclear granulocytes, disturbed crypt architecture and epithelial damage with mucusd ve aftert
j(ihpapffrp
3
WSRw<a
4
F
(eMihar(6snc4ios(
wsitdg
epletion (pointed) and a crypt abscess (pointed) which dramatically improhe epithelial damage with a quite normal goblet cell population.
ects (matched for sex and age) with irritable bowel syndromeIBS), in whom colonoscopy had been performed for abdom-nal pain, were used as control. In IBS biopsy samples showedistologically normal epithelium and mucosa. For the pur-ose of the study, all biopsy specimens were assessed bysingle expert gastrointestinal pathologist (E.I.) without
rior knowledge of the time of treatment. Biopsy specimensrom the same patient before and after treatment, both takenrom the most severely inflamed areas, were compared. Theesults were also compared with those obtained from IBSatients.
. Statistical analysis
Data analysis was performed by using Student’s t-test,ilcoxon’s test and Mann–Whitney U test as appropriate.
ample size was calculated, as suggested by Sokal andohlf [18]. A possible correlation among different findingsas evaluated according to the Spearman’s test. A p value0.05 was considered as indicative of statistical signific-nce.
. Results
Nine patients were enrolled between March 2005 andebruary 2006. The median Mayo score at entry was 8
i(ws
therapy (B) with regression of the inflammatory infiltrate and restoring of
range: 7–11). All patients had a moderate or severe dis-ase activity at flexible sigmoidoscopy (median endoscopicayo subscore: 3, range 2–3). Similarly, a moderate to severe
nflammatory activity was detected in all biopsies (medianistologic score: 3, range 2–3) (Fig. 1). This was associ-ted with a high number of neutrophils (median: 17.2/mm2;ange 7.8–20), and a moderate number of lymphocytesmedian: 7.5/mm2; range 4.5–14) and plasma cells (median:.3/mm2; range 5.4–12.4). Such values were statisticallyignificantly higher as compared to IBS controls (medianeutrophils: 1.6/mm2; range: 0–3; p < 0.01; median lympho-ytes: 4.1/mm2; range: 3–5.5; p = 0.04; median plasma cells:.3/mm2: 2.7–4.8; p < 0.01). The increased histologic activ-ty in UC patients was also related with a high expressionf TNF-� in the colonica mucosa. Immunohistochemicaltaining showed a TNF-� score of 40% (range: 29–58.3%)Fig. 2).
All patients received the scheduled doses of infliximab ateeks 0, 2 and 6, respectively. At week 10, a clinical remis-
ion was achieved in two (22.2%) patients, a clinical responsen further four (44.5%), whilst no response was observed inhe remaining three (33.3%) cases. Characteristics of respon-er (remission or response) and non-responder patients areiven in Table 1. As shown in Fig. 3, Mayo score significantly
mproved in the responders as compared to baseline valuesmedian: 3.5; range 1–5; p = 0.002), but not in the patientsho did not respond (median: 7; range: 6–9; p = 0.6). Flexible
igmoidoscopy at week 10 showed a significant improvement
814 C. Hassan et al. / Digestive and Liver Disease 39 (2007) 811–817
Fig. 2. Immunohistochemical changes in colonic TNF-� expression beforeand after infliximab therapy (immunohistochemical stains; original magni-fication 400×). A high TNF-� level may be depicted before therapy withpositive cells stained red (pointed) and negative ones stained blue (haema-toxylin) (A), with a profound down-regulation after treatment (B).
Table 1Demographic and clinical data at entry
Responder (n = 6) Non-responder (n = 3)
Age (year), mean ± S.D. 48 ± 17 46 ± 15Sex (M/F) 4/2 2/1Duration of disease,
mean ± S.D.7.5 ± 2.2 5.3 ± 1.5
Mayo score, median (range) 9 (7–11) 8 (7–8)
Colonic area involvedLeft-side 2 1Extensive 4 2
Endoscopic Mayo subscore,median (range)
3 (2–3) 2 (2–3)
Histologic score, median(range)
2.75 (2–3) 3 (2–3)
Concomitant medicationsSul-fasalazine/mesalamine
4 2
Corticosteroids 5 3Azathioprine 5 2
Fa
ortIarsiw(trTit
ig. 3. Changes of clinical (A), endoscopic (B) and histological scores (C)fter infliximab treatment in responder (�) and non-responder (�) patients.
f the Mayo subscore in the responder patients (median: 1;ange: 0–2; p = 0.004), whilst no change was observed inhe non-responder group (median: 3; range: 2–3; p = 0.6).n detail, mucosal healing occurred in five responder patientsnd in none of the failure patients. After therapy, a dramaticegression of the inflammatory activity was observed. Ashown in Fig. 3, the histological score significantly decreasedn the responder group (median: 0.5; range: 0–1.5; p = 0.002),hilst no change was observed in those non-responders
median: 3; range: 2–3). Among responders, mucosal archi-ecture returned to normal in two cases, whilst an incomplete
ecovery was observed in the remaining cases. As shown inable 2, the histological improvement was mainly due to an
mpressive regression of the number of neutrophils associatedo the restoration of a normal crypt architecture and mucus
C. Hassan et al. / Digestive and Liver
Fa
cb1IotTdnsrtea(
Tsspp
5
d
wmfiaraiiehprdaIraWAs
hshspianctrtpmspumoFiadocafioIa
ig. 4. Changes of TNF-� score (A) and absolute number of neutrophils (B)fter infliximab treatment in responders (�) and non-responders (�).
ontent in epithelial cells (see also Fig. 4). However, the num-er of neutrophils after therapy (median: 4.9/mm2; range:.9–7.1) was still significantly higher than that observed inBS controls (median: 1.6/mm2; range: 0–3; p = 0.04). On thether hand, no significant reduction of mononuclear cell infil-ration was observed both in responders and non-responders.he overall number of TNF-� positive cells dramaticallyecreased after infliximab therapy in responders, with no sig-ificant change observed in non-responders (Table 2). TNF-�core at week 10 was not only significantly reduced in theesponder patients, but it was also similar to that observed inhe IBS controls (median: 8%; range 4–12%; p = 0.2). Inter-stingly, the number of positive cells was more than halveds compared to pre-treatment values in all responder patientsFig. 4).
Grouping altogether pre- and post-treatment values,NF-� score appeared to be correlated with the Mayocore (r = 0.78; p = 0.0004), the endoscopic Mayo sub-core (r = 0.80; p = 0.0003), the histologic score (r = 0.59;= 0.001), and with the number of neutrophils (r = 0.72;= 0.002).
. Discussion
Infliximab seems to trigger a long-lasting profoundown-regulation of mucosal TNF-� expression in patients
sosp
Disease 39 (2007) 811–817 815
ith moderate-to-severe UC. Indeed, a dramatic decline ofucosal TNF-� levels was detected after 10 weeks after therst infusion in our study. Moreover, such down-regulationppeared to be intimately associated with an impressiveegression of the active inflammatory process. In particular,n almost complete disappearance of neutrophils occurredn our population. Importantly, such an immunohistochem-cal effect was strongly associated with both clinical andndoscopic improvement. Notably, TNF-� down-regulation,istological regression, and endoscopic improvement wereurely confined to those patients achieving a clinicalesponse/remission. In fact, we did not detect TNF-�own-regulation or significant reduction of the histologicalctivity in patients who did not respond to infliximab therapy.t has been recently suggested that therapeutic failure inheumatoid arthritis may be related to the polymorphismt position -308 in the promoter of TNF-� gene [19].
hether a poor response in UC is also associated with an/A genotype in this locus needs to be assessed by further
tudies.UC pathogenesis has been widely related with a type 2
elper T cell response, contrary to CD, which has been con-idered a type 1 helper T cell-related disease [20]. Our dataeavily questions the validity and the clinical usefulness ofuch a classification. TNF-�, a type 1 cytokine, appeared tolay a major role in UC pathogenesis in our study. Not onlyt was over-expressed in patients with an active disease butlso its down-regulation following a blocking antibody wasecessary to achieve a histological regression and a clini-al and endoscopic response. Such evidence tightly mirrorshe high efficacy of infliximab in inducing and maintainingemission in moderate-to-severe UC patients in two large con-rolled trials [10]. To address this discrepancy, it should beointed out that the identification in vitro of two types ofurine helper T cell clones – type 1 and 2 – has never been
upported by their detection in human diseases, whilst theattern of cytokine produced – either type 1 or 2 – has beensed to classify a human disorder as a type 1 or 2 T helperediated [21]. However, the production of only one type
f cytokines has not been described in any human disease.or example, IL-4 – a type 2 cytokine – has been described
n CD (reported as being a Th-1 disease), whilst TNF-�,type 1 cytokine, has been reported in UC as well, so far
escribed as a Th-2 disease [13,6–8]. Similarly, the efficacyf the biological drugs does not seem to be necessarily asso-iated with such a classification. For example, a therapeuticttempt with IL-10 – an inhibitory type 2 cytokine – hasailed in UC [22]. To further confound our confidence, inflix-mab treatment has been associated with a down-regulationf type 2 cytokines in CD and rheumatoid arthritis, such asL-4 [13]. Taking all these findings into account, it seemsdvisable not to prevent a potentially effective and life-
aving biologic therapy only on the basis of our knowledgef the pathogenesis of immune-mediated disorders, whichhould, on the other hand, be reserved mainly for researchurposes.
816 C. Hassan et al. / Digestive and Liver Disease 39 (2007) 811–817
Table 2TNF-� scores and number of inflammatory cells per mm2 at baseline (week 0) and after infliximab treatment (week 10) in mucosal biopsies of responders andnot responders
• A profound down-regulation of TNF-alphaoccurs after infliximab therapy in ulcerativecolitis.
• TNF-alpha down-regulation is confined onlyto those patients who achieved a clinicalresponse.
• Such down-regulation is associated with adramatic regression of the inflammatory pro-cess, and with an endoscopic and clinicalimprovement.
Research agenda
• Genetic factors responsible for anti-TNF-�therapeutic failure in non-responder patients.
• Immunological mechanisms responsible forTNF-� recurrence in responder patients.
• Immunohistochemical factors which maypredict therapeutic response, similarly to
CN
R
a All values are expressed as median (range).b Only for responder cases.
The immunohistological response to infliximab observedn our UC patients closely resembled the same effect previ-usly described in CD by Baert et al. [13], who detected arofound down-regulation of TNF-� associated with a dra-atic regression of the inflammatory activity by 4 weeks aftersingle infusion of infliximab. Our study reported practically
he same scenario in a different disease after a longer time.uggestively, we also described a similar down-regulation 1ear after the first infliximab infusion in a case of segmentalolitis associated with diverticula [23]. All these observa-ions seem to indicate that the long-term effect of infliximabn inflammatory bowel diseases is mainly related to the per-istence of the TNF-� down-regulation and to the relatednhibition of the inflammatory process. However, the rea-ons of this long-lasting effect are poorly understood. It haseen recently suggested that one of the possible mecha-isms is related to the apoptosis of TNF-� expressing cellsesponsible for mucosal inflammation, such as monocytesnd lymphocytes, via antibody-dependent cell-mediated andomplement-dependent cytotoxicity [24,25]. Nevertheless,e failed to detect possible signs of cellular lysis, such asuclear dust or an increase of pigment-loaded macrophages,imilarly to what previously reported by Baert [13]. Alter-atively, it could be speculated that the therapeutic effects related to an inhibition of the migration of inflamma-ory cells from the systemic circulation through the bowelall into the lamina propria of the intestinal mucosa. Such
ell trafficking is mainly regulated by a variety of adhesionolecules among which ICAM-1 and its ligand LFA-1, asell as syndecan-1, have been shown to play an important role
n inflammatory bowel diseases [26,27]. However, a reduc-ion of these molecules has not been described neither inheumatoid arthritis nor in CD following infliximab infusions13,28].
In conclusion, our study has shown a profound down-egulation of TNF-� in moderate-to-severe UC afternfliximab, providing a compelling rationale for such a treat-
ent. Infliximab appeared to be able to trigger a “reset ofhe immunostat” in a type 2 helper T cell-related disorder,
ontrary to what previously believed [29]. A further evidencehat the more we struggle to unravel the inflammatory puzzle,he more puzzling the pathogenetic enigma of inflammatoryowel diseases appears to be.
oestrogen receptors in breast cancer.
onflict of interest statementone declared.
eferences
[1] Reimund JM, Wittersheim C, Dumont S, Muller CD, Kenney JS, Bau-mann R, et al. Increased production of tumour necrosis factor-alphainterleukin-1 beta, and interleukin-6 by morphologically normal intesti-nal biopsies from patients with Crohn’s disease. Gut 1996;39:684–9.
[2] Monteleone G, Monteleone I, Fina D, Vavassori P, Del Vecchio BlancoG, Caruso R, et al. Interleukin-21 enhances T-helper cell type I signaling
and interferon-gamma production in Crohn’s disease. Gastroenterology2005;128:687–94.
[3] Fuss IJ, Neurath M, Boirivant M, Klein JS, de la Motte C, StrongSA, et al. Disparate CD4+ lamina propria (LP) lymphokine secretion
profiles in inflammatory bowel disease. Crohn’s disease LP cells man-ifest increased secretion of IFN-gamma, whereas ulcerative colitis LPcells manifest increased secretion of IL-5. J Immunol 1996;157:1261–70.
[4] Plevy SE, Landers CJ, Prehn J, Carramanzana NM, Deem RL, ShealyD, et al. A role for TNF-alpha and mucosal T helper-1 cytokines in thepathogenesis of Crohn’s disease. J Immunol 1997;159:6276–82.
[5] Berner B, Akca D, Jung T, Muller GA, Reuss-Borst MA. Analysisof Th1 and Th2 cytokines expressing CD4+ and CD8+ T cells inrheumatoid arthritis by flow cytometry. J Rheumatol 2000;27:1128–35.
[6] Murch SH, Lamkin VA, Savage MO, Walker-Smith JA, MacDonaldTT. Serum concentrations of tumour necrosis factor alpha in childhoodchronic inflammatory bowel disease. Gut 1991;32:913–7.
[7] Murch SH, Braegger CP, Walker-Smith JA, MacDonald TT. Locationof tumour necrosis factor alpha by immunohistochemistry in chronicinflammatory bowel disease. Gut 1993;34:1705–9.
[8] Braegger CP, Nicholls S, Murch SH, Stephens S, MacDonald TT.Tumour necrosis factor alpha in stool as a marker of intestinal inflam-mation. Lancet 1992;339:89–91.
[9] Knight DM, Trinh H, Le J, Siegel S, Shealy D, McDonough M, et al.Construction and initial characterization of a mouse-human chimericanti-TNF-antibody. Mol Immunol 1993;30:1443–53.
10] Rutgeerts P, Sandborn WJ, Feagan BG, Reinisch W, Olson A, JohannsJ, et al. Infliximab for induction and maintenance therapy for ulcerativecolitis. N Engl J Med 2005;353:2462–76.
11] Hanauer SB, Feagan BG, Lichtenstein GR, Mayer LF, Schreiber S,Colombel JF, et al. Maintenance infliximab for Crohn’s disease: theACCENT I randomised trial. Lancet 2002;359:1541–9.
12] Sands BE, Anderson FH, Bernstein CN, Chey WY, Feagan BG, Fedo-rak RN, et al. Infliximab maintenance therapy for fistulising Crohn’sdisease. N Engl J Med 2004;350:876–85.
13] Baert FJ, D’Haens GR, Peeters M, Hiele MI, Schaible TF, ShealyD, et al. Tumor necrosis factor alpha antibody (infliximab) therapyprofoundly down-regulates the inflammation in Crohn’s ileocolitis.Gastroenterology 1999;116:22–8.
14] Ulfgren AK, Andersson U, Engstrom M, Klareskog L, Maini RN, Tay-lor PC. Systemic anti-tumor necrosis factor alpha therapy in rheumatoidarthritis down-regulates synovial tumor necrosis factor alpha synthesis.Arthritis Rheum 2000;43:2391–6.
15] Schroeder KW, Tremaine WJ, Ilstrup DM. Coated oral 5-aminosalicylicacid therapy for mildly to moderately active ulcerative colitis: a ran-domized study. N Engl J Med 1987;317:1625–9.
17] Ierardi E, Monno RA, Gentile A, Francavilla R, Burattini O, MarangiS, et al. Helicobacter Heilmannii gastritis: a histological and immuno-histochemical trait. J Clin Pathol 2001;54:774–7.
19] Seitz M, Wirthmuller U, Moller B, Villiger PM. The -308 tumour necro-sis factor-{alpha} gene polymorphism predicts therapeutic responseto TNF{alpha}-blockers in rheumatoid arthritis and spondyloarthritispatients. Rheumatology 2007;46:93–6.
20] Targan SR, Karp LC. Defects in mucosal immunity leading to ulcerativecolitis. Immunol Rev 2005;206:296–305.
21] Mosmann TR, Cherwinski H, Bond MW, Giedlin MA, Coffman RL.Two types of murine helper T cell clone. I. Definition according toprofiles of lymphokine activities and secreted proteins. 1986. J Immunol2005;175:5–14.
22] Schreiber S, Fedorak RN, Nielsen OH, Wild G, Williams CN, Niko-laus S, et al. Safety and efficacy of recombinant human interleukin 10in chronic active Crohn’s disease. Crohn’s Disease IL-10 CooperativeStudy Group. Gastroenterology 2000;119:1461–72.
23] Hassan C, Zullo A, Ierardi E, Burattini O, De Francesco V, Morini S.Tumour necrosis factor alpha downregulation and therapeutic responseto infliximab in a case of segmental colitis associated with diverticula.Gut 2006;55:589–90.
24] Lugering A, Schmidt M, Lugering N, Pauels HG, Domschke W,Kucharzik T. Infliximab induces apoptosis in monocytes from patientswith chronic active Crohn’s disease by using a caspase-dependent path-way. Gastroenterology 2001;121:1145–57.
25] ten Hove T, van Montfrans C, Peppelenbosch MP, van Deventer SJ.Infliximab treatment induces apoptosis of lamina propria T lympho-cytes in Crohn’s disease. Gut 2002;50:206–11.
26] Westermann B, Engelhardt B, Hoffmann J. Migration of T-cells invivo: molecular mechanisms and clinical implications. Ann Intern Med2001;135:279–95.
27] Principi M, Day R, Marangi S, Burattini O, De Francesco V, IngrossoM, et al. Differential immunohistochemical expression of syndecan-1 and tumor necrosis factor alpha in colonic mucosa of patients withCrohn’s disease. Immunopharmacol Immunotoxicol 2006;28:185–95.
28] Macias I, Garcia-Perez S, Ruiz-Tudela M, Medina F, Chozas N, Giron-Gonzalez JA. Modification of pro- and antiinflammatory cytokinesand vascular-related molecules by tumor necrosis factor-a blockade
in patients with rheumatoid arthritis. J Rheumatol 2005;32:2102–8.
29] Mizoguchi E, Mizoguchi A, Takedatsu H, Cario E, de Jong YP, OoiCJ, et al. Role of tumor necrosis factor receptor 2 (TNFR2) in colonicepithelial hyperplasia and chronic intestinal inflammation in mice. Gas-troenterology 2002;122:134–44.
