Detection of Pro- and Anti-Inflammatory Cytokines in Stools of Patientswith Inflammatory Bowel Disease

T. SAIKI, K. MITSUYAMA, A. TOYONAGA, H. ISHIDA & K. TANIKAWASecond Dept. of Medicine, Kurume University, School of Medicine, Kurume, and National UtanoHospital, Kyoto, Japan

Saiki T, Mitsuyama K, Toyonaga A, Ishida H, Tanikawa K. Detection of pro- and anti-inflammatorycytokines in stools of patients with inflammatory bowel disease. Scand J Gastroenterol 1998;33:616–622.

Background:Cytokines play a predominant role in immune and inflammatory reactions in inflammatorybowel disease. Any cytokine that is produced locally as a result of gut inflammation may leak into thebowel lumen and appear in the stools. We examined the usefulness of determining cytokine profiles in thestools of patients with ulcerative colitis or Crohn’s disease.Methods:Cytokine concentrations in stoolextracts were measured in 36 patients with ulcerative colitis, 32 patients with Crohn’s disease, 9 controlswith inflammatory disease, and 18 normal controls by means of enzyme-linked immunosorbent assays.Results:Stool concentrations of interleukin-1b and interleukin-1 receptor antagonist in patients withactive inflammatory bowel disease increased significantly and correlated with various inflammatoryfactors and stool concentrations of polymorphonuclear cell elastase. The ratio of interleukin-1 receptorantagonist to interleukin-1b in active disease was reduced significantly compared with that in inactivedisease or in normal controls. Paired analysis showed a decrease in tumor necrosis factor-a andinterleukin-1b and interleukin-1 receptor antagonist and an increase in interleukin-4 and interleukin-10concentrations after the resolution of disease exacerbation.Conclusions:Measurement of cytokines instools may be a useful and noninvasive means of understanding pathophysiology and clinical monitoringin inflammatory bowel disease.

Key words:Cytokine; inflammatory bowel disease; stool

Keiichi Mitsuyama, M.D., Second Dept. of Medicine, Kurume University School of Medicine, 67 Asahi-machi, Kurume, Fukuoka 830, Japan (fax:�81 942-34-2623)

Although complete understanding of the role of cytokines inthe pathophysiology of inflammatory bowel disease (IBD) hasnot been achieved, it is increasingly evident that thesemolecules in aggregate play a predominant role in immuneand inflammatory reactions in this disorder (1–3). Whereasthe concentration of cytokines in the intestine itself isimportant for the characterization of cytokine profiles inIBD (4–8), it is of course not suitable for practical use. It isalso not known whether cytokines present in the peripheralcirculation mirror intestinal inflammation. Therefore, weasked whether the stools of IBD patients contain increasedamounts of cytokines and, if so, whether measurement ofstool cytokines would be a reliable and noninvasive means ofunderstanding the disease status in the gut. Further, pre-liminary reports have described significant amelioration ofdisease activity in patients with Crohn’s disease (CD) aftersystemic administration of anti-tumor necrosis factor (TNF)-a

monoclonal antibodies (9) and in patients with ulcerativecolitis (UC) after intracolonic administration of interleukin(IL)-10 (10). Therefore, the monitoring of cytokine profilesmay be more relevant for evaluating the efficacy ofimmunotherapy. However, except for two reports on apediatric population showing a significant increase in TNFa

(11) and IL-6 (12) during active disease, little is known aboutcytokine profiles in the stools of patients with IBD.

In the present study we identified and quantitated cytokineswith pro- and anti-inflammatory properties in stools of IBDpatients, using specific and sensitive enzyme-linked immu-nosorbent assays (ELISAs), and examined how the cytokineswere related to clinical variables.

MATERIALS AND METHODS

EthicsThis project was performed in accordance with the Helsinki

Declaration and was approved by the local medical ethicscommittee. Informed consent was obtained from everypatient.

SubjectsThirty-six patients with UC and 32 patients with CD were

investigated. The diagnoses were based on characteristicclinical, endoscopic, radiologic, and histologic features.

Patients with UC.There were 23 men and 13 women with amedian age of 32 years (range, 17–65 years). In terms ofdisease distribution, 19 patients had pancolitis, eight patientshad left colon involvement, and 9 patients had disease limitedto the rectum. Disease activity in each patient was analyzed inaccordance with the criteria of Truelove & Witts (13).Nineteen patients had active disease (5, mild; 12, moderate;

Scan

d J

Gas

troe

nter

ol D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsita

t Aut

onom

a B

arce

lona

on

10/2

7/14

For

pers

onal

use

onl

y.

2, severe) and 17 patients had inactive disease. At the time ofstudy 13 patients had taken sulphasalazine only (5 active, 8inactive); 7 patients, both corticosteroid and sulphasalazine (2active, 5 inactive); 2 patients, corticosteroid only (2 active, 0inactive); and 14 patients, no specific treatment (10 active, 4inactive). Whenever possible, endoscopic findings weregraded from 1 to 4 in accordance with the severity ofinflammation, using the criteria of Matts (14).

Patients with CD.There were 20 men and 12 women, witha median age of 34 years (range, 17–67 years). In 14 patientsthe disease affected both the ileum and the colon, in 5 patientsthe colon only, and in 13 patients the ileum only. Diseaseactivity was assessed on the basis of the score of theInternational Organization for the Study of InflammatoryBowel disease (IOIBD) (15). A score of 1 or less was defined

as corresponding to inactive disease, and a score of 2 or morecorresponded to more active disease. In this group 18 patientshad active and 14 patients inactive disease. Ten patients hadtaken sulphasalazine only (5 active, 5 inactive); 8 patients,both corticosteroid and sulphasalazine (4 active, 4 inactive);and 14 patients, no specific treatment (9 active, 5 inactive).

Control subjects.Eighteen healthy, age-matched subjectsserved as normal controls. Nine patients with other colitides(seven infectious and two ischemic colitis) were examined asinflammatory disease controls.

SamplingPatients were instructed to defecate directly into a poly-

styrene container (diameter, 11.5 cm; depth, 7.5 cm). Thestool extraction procedure for cytokine measurement has been

Fig. 1. Stool concentrations of polymorphonuclear cell elastase (PME), interleukin (IL)-1b, IL-1 receptorantagonist (IL-1ra), tumor necrosis factor (TNF)-a, IL-4, and IL-10, and the ratio of IL-1ra to IL-1b innormal subjects (A,n = 18) and in patients with active (B,n = 19) and inactive (C,n = 17) ulcerativecolitis, active (D,n = 18) and inactive (E,n = 14) Crohn’s disease, and inflammatory controls (F,n = 9).n = number of patients. The data represent the mean� standard error of the mean.

Stool Cytokine Levels in IBD 617

Scan

d J

Gas

troe

nter

ol D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsita

t Aut

onom

a B

arce

lona

on

10/2

7/14

For

pers

onal

use

onl

y.

described previously (16, 17). In brief, stool was weighed,diluted 1:2 in phosphate-buffered saline (pH 7.2) containingsoy trypsin inhibitor (1 mg/ml; Wako Pure Chemical Indus-tries, Ltd., Osaka, Japan) and phenylmethylsulfonyl fluoride(1 mg/ml; Wako Pure Chemical Industries, Ltd.) and cen-trifuged at 10,000g at 4°C for 15 min. Supernatants werecollected and kept frozen atÿ80°C until use. A blood samplewas also obtained to determine various laboratory variables.

Determination of stool samplesCytokines were measured in the stools, using sensitive

ELISAs for IL-1b, TNFa (detection limit,<0.1 pg/ml for IL-1b and<0.18 pg/ml for TNFa; R&D Systems Inc., Minnea-polis, Minn., USA), IL-1 receptor antagonist (IL-1ra)(<22.0 pg/ml, Amersham International plc, Biotrak, UK),IL-4 (<2.1 pg/ml, SRL Inc., Tokyo, Japan) (18), and IL-10(<2.0 pg/ml, Sumitomo Metal Bio-Science Inc., Kanagawa,Japan) (19). A stool marker for disease activity, polymorpho-nuclear cell elastase, was determined with ELISA (<11.0mg/l, Sanwa Kagaku Kenkyusho Co., Ltd., Nagoya, Japan) (20).

Validity testingThe reliability of stool cytokine measurement was tested by

making the following observations: a) The recovery rate fromstool extraction was determined by mixing known quantitiesof cytokines with stools before grinding. Recovery ofexogenously administrated cytokines was uniformly>90%.b) Sample titration gave detectable levels of cytokines thatalways followed the standard curve. c) The variation ofcytokine concentration in stool samples obtained from thesame individual on different days (within 3 days) was<12.0% (n = 5).

Determination of blood samplesRoutine clinical laboratory methods were used for addi-

tional assessment of acute-phase indicators, including theleukocyte count and erythrocyte sedimentation rate. Serum C-reactive protein concentrations (normal range<350 ng/ml)were measured with laser nephelometry (NA latex CRP kit;Hoechst Japan, Tokyo, Japan).

Statistical analysisThe means of multiple groups were compared by using

Bonferroni’s test after analysis of variance. Student’st,Kruskal–Wallis, and Spearman rank correlation were alsoused for statistical analyses. Differences were consideredstatistically significant for anyP< 0.05.

RESULTS

Individual stool cytokine concentrations are shown in Fig. 1.Cytokines were detected in every subject. Stool concentra-tions of IL-1b and IL-1ra were significantly higher in patientswith active IBD than in patients with inactive disease andnormal controls. There was no significant difference in IL-1b T

able

I.C

orre

latio

nco

effic

ient

san

dsi

gnifi

canc

eof

diffe

renc

esbe

twee

nst

oolc

ytok

ine

conc

entr

atio

nsan

dot

her

labo

rato

ryva

riabl

esin

patie

nts

with

infla

mm

ator

ybo

wel

dise

ase

Ulc

erat

ive

colit

isC

rohn

’sdi

seas

e

IL-1

b

IL-1

raT

NF

-a

IL-4

IL-1

0IL

-1

b

IL-1

raT

NF

a

IL-4

IL-1

0

Leuk

ocyt

eco

unt

ÿ

0.03

(31)

ÿ

0.05

(31)

ÿ

0.15

(31)

ÿ

0.19

(31)

0.10

(31)

0.34

†(2

8)0.

33(2

8)

ÿ

0.31

(27)

0.09

(28)

ÿ

0.06

(28)

1-h

ES

R0.

35‡

(28)

0.48

‡(2

8)0.

23(2

8)

ÿ

0.10

(28)

ÿ

0.06

(28)

0.55

‡(2

5)0.

36(2

5)0.

13(2

4)0.

23(2

5)

ÿ

0.04

(25)

C-r

eact

ive

prot

ein

0.44

†(3

5)0.

45*

(35)

0.34

(35)

ÿ

0.06

(35)ÿ

0.08

(35)

0.42

†(3

0)0.

05(3

0)0.

14(2

9)

ÿ

0.14

(30)

ÿ

0.17

(30)

PM

E0.

84‡

(35)

0.64

‡(3

5)0.

51*

(35)

ÿ

0.16

(35)ÿ

0.18

(35)

0.07

‡(3

0)0.

09*

(30)

ÿ

0.15

(30)

ÿ

0.11

(30)

ÿ

0.10

(30)

IL-1

b

0.51

‡(3

6)0.

46(3

6)

ÿ

0.11

(36)

ÿ

0.23

*(3

6)0.

21†

(32)

ÿ

0.06

(31)

ÿ

0.06

(32)

ÿ

0.06

(32)

IL-1

ra0.

51‡

(36)

0.35

*(3

6)

ÿ

0.20

(36)

ÿ

0.15

(36)

0.21

†(3

2)

ÿ

0.12

(31)

0.04

(32)

ÿ

0.08

(32)

TN

F

a

0.46

(36)

0.35

*(3

6)

ÿ

0.03

(36)

0.04

(36)

ÿ

0.06

(31)

ÿ

0.12

(31)

0.04

(31)

0.22

(31)

IL-4

ÿ

0.11

(36)

ÿ

0.20

(36)

ÿ

0.03

(36)

0.04

(36)

ÿ

0.06

(32)

0.04

(32)

0.04

(31)

0.41

†(3

2)IL

-10

ÿ

0.23

*(3

6)

ÿ

0.15

(36)

0.04

(36)

0.04

(36)

ÿ

0.06

(32)

ÿ

0.08

(32)

0.22

(31)

0.41

†(3

2)

IL=

inte

rleuk

in;

TN

F=

tum

orne

cros

isfa

ctor

;E

SR

=er

ythr

ocyt

ese

dim

enta

tion

rate

;P

ME

=po

lym

orph

onuc

lear

cell

elas

tase

;IL

-1ra

=IL

-1re

cept

oran

tago

nist

.F

igur

esin

pare

nthe

ses

are

num

bers

ofpa

tient

s.*

P

<

0.05

;†

P

<

0.00

5;‡P

<

0.00

1.

618 T. Saiki et al.

Scan

d J

Gas

troe

nter

ol D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsita

t Aut

onom

a B

arce

lona

on

10/2

7/14

For

pers

onal

use

onl

y.

and IL-1ra concentrations between active UC and CD. Theratio of IL-1ra to IL-1b in patients with active IBD wassignificantly lower than that in inactive disease or in normalcontrols. Stool concentrations of IL-4 and IL-10 in activedisease were comparable to those of patients with inactivedisease or controls.

There was significant correlation of stool IL-1b and IL-1rawith laboratory variables or stool concentrations of poly-morphonuclear cell elastase in both diseases (Table I).Further, stool IL-1b correlated significantly with the endo-scopic grade of UC (P< 0.05) (data not shown). Statisticalassociation was also observed between IL-4 and IL-10 in thestool of CD patients and between IL-1b and IL-1ra in both UCand CD patients (Table I).

Stool cytokine concentrations in the active and ensuinginactive phase were compared in 7 patients with UC and 5

with CD. As shown in Fig. 2, IL-1b, IL-1ra, and TNFadecreased significantly after clinical resolution. Interestingly,IL-4, IL-10, and the ratio of IL-1ra to IL-1b were significantlyincreased after the resolution of disease exacerbation.

Stool cytokine concentrations were analyzed in relation tothe disease location and medical treatment. Concentrations ofall five cytokines were independent of the disease location inboth forms of IBD (Table II). In addition, no significantdifferences in cytokine concentration were evident betweenactive disease patients with steroid therapy and those without,except for IL-4 in UC (Table III).

DISCUSSION

There are many reports on cytokine profiles in the intestine ofpatients with IBD (5–8, 21–29). In the present investigation,

Fig. 2. Serial levels of interleukin (IL)-1b, IL-1 receptor antagonist (IL-1ra), tumor necrosis factor (TNF)-a, IL-4, IL-10, and the ratio of IL-1ra to IL-1b in stools of patients with inflammatory bowel disease.Paired stool samples in both active (A) and inactive (I) phases were obtained from seven patients withulcerative colitis (UC) and five with Crohn’s disease (CD).

Stool Cytokine Levels in IBD 619

Scan

d J

Gas

troe

nter

ol D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsita

t Aut

onom

a B

arce

lona

on

10/2

7/14

For

pers

onal

use

onl

y.

we focused on the appearance of cytokines in the stool ofthese patients.

We found increases in both IL-1 and its natural antagonistIL-1ra (30) in the stools of patients with active IBD and lowlevels in inactive disease and controls. This observation issimilar to previous findings of increased tissue levels of IL-1and IL-1ra in patients with IBD when using either wholebiopsy specimens (24) or isolated lamina propria mono-nuclear cells (25). Moreover, we found a significant correla-tion of both IL-1 and IL-1ra with various clinical indicators ofdisease activities or severities. Although the design of thisstudy did not enable further comparison of these twocytokines with the degree of disease activity, the presentresults suggest that stool IL-1 and IL-1ra are simple andrelevant markers of IBD activity.

There is a general trend for the ratio of IL-1ra to IL-1 to bedecreased in whole biopsy specimens (24) or cell extracts (25)from UC and CD patients when compared with controls. Asimilar trend was found in stool samples reported here. Theseobservations suggest that inflammation could result fromabnormal proportions of IL-1 and IL-1ra secondary to arelative excess of the former or insufficiency of the latter.

TNFa is a key inflammatory cytokine that shares severalproperties with IL-1 (31). Previous study in a pediatricpopulation showed an increased TNFa concentration in thestools of patients with active IBD. Somewhat surprisingly, wecould not find statistically significant differences in TNFa

levels between the active and inactive CD groups. Tissuelevels of TNFa have been measured in several studies with

conflicting results (5, 6, 26). It is intriguing that TNFa is notconsistently increased in IBD patients in a manner similar tothat of the other pro-inflammatory cytokines, IL-1, IL-6, andIL-8. Differences in patient selection, tissue sampling, anddiverse methods may in part explain these conflictingobservations.

In contrast to other cytokines, the opposite situation arisesfor IL-4 (32, 33) and IL-10 (34, 35), both of which have thecapacity to inhibit pro-inflammatory cytokines while stimu-lating IL-1ra. In a paired analysis of the same individual, stoolconcentrations of IL-4 and IL-10 have been shown to increaseafter the clinical resolution of disease exacerbation. Also,there was a significant correlation between stool concentra-tions of IL-4 and IL-10, especially in CD. The exact meaningof the delay in the increase in these two anti-inflammatorycytokines is unclear. However, it is conceivable that IL-4 andIL-10 may contribute to systemic down-regulation of theacute inflammatory process. Clinically, serial measurementsin the same individual may help determine disease prognosisand confirm treatment efficacy. At present, conflicting resultshave been observed in tissue levels of both IL-4 and IL-10 inrelation to disease activities (8, 27–29). Ideally, bothintestinal tissue and stool samples should be evaluatedsimultaneously in the same patient.

It is well established that currently available drugs,particularly corticosteroids, attenuate inflammation in thebowel by suppressing the production of cytokines such as IL-1 and TNFa (36, 37). However, we could not find anydifference in cytokine levels between patients with and

Table II. Stool cytokine concentrations and disease location in patients with active inflammatory bowel disease

Disease location No. of patients IL-1b (log pg/g) IL-1ra (log pg/g) TNFa (log pg/g) IL-4 (log pg/g) IL-10 (log pg/g)

Ulcerative colitisProctitis 3 4.76� 0.33 5.04� 0.44 0.95� 0.38 2.06� 0.32 2.39� 0.24Left-sided colitis 5 4.56� 0.27 5.26� 0.17 1.24� 0.36 2.15� 0.23 2.46� 0.36Pancolitis 11 4.64� 0.42 5.14� 0.27 1.94� 0.18 2.56� 0.16 2.51� 0.27

Crohn’s diseaseIleitis 7 2.89� 0.34 4.30� 0.36 0.92� 0.19 2.17� 0.27 2.44� 0.15Colitis 3 5.08� 0.32 5.03� 0.17 1.30� 0.33 2.73� 0.69 3.14� 0.55Ileocolitis 8 3.61� 0.43 4.26� 0.24 1.01� 0.25 2.04� 0.14 2.21� 0.15

IL = interleukin; TNF = tumor necrosis factor; IL-1ra = IL-1 receptor antagonist. Cytokine levels were compared using Bonferroni’s testafter analysis of variance with logarithmically transformed data. The data represent mean� standard error of the mean.

Table III. Stool cytokine concentrations and medical treatment in patients with active inflammatory bowel disease

Medication No. of patients IL-1b (log pg/g) IL-1ra (log pg/g) TNFa (log pg/g) IL-4 (log pg/g) IL-10 (log pg/g)

Ulcerative colitisNone 10 4.97� 0.36 5.39� 0.14 1.85� 0.17 2.44� 0.09 2.44� 0.24Steroid* 4 4.82� 0.27 5.33� 0.30 1.27� 0.48 1.82� 0.25** 2.71� 0.41

Crohn’s diseaseNone 9 3.01� 0.29 4.19� 0.30 0.87� 0.22 2.07� 0.22 2.36� 0.12Steroid* 4 3.93� 0.59 4.58� 0.33 1.37� 0.03 2.63� 0.48 2.95� 0.41

IL = interleukin; TNF = tumor necrosis factor; IL-1ra = IL-1 receptor antagonist. Cytokine levels were compared using Student’st test withlogarithmically transformed data. The data represent mean� standard error of the mean.

* Patients treated with steroid alone and both steroid and sulphasalazine.** P< 0.05 versus ulcerative colitis patients without specific medication.

620 T. Saiki et al.

Scan

d J

Gas

troe

nter

ol D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsita

t Aut

onom

a B

arce

lona

on

10/2

7/14

For

pers

onal

use

onl

y.

without steroid therapy, except for IL-4 in UC. Because of thesmall size of each subgroup, no final conclusion could beextrapolated with regard to the influence of steroids on stoolcytokine levels. Further studies in a larger population areneeded.

The question arose as to how cytokines appear in the stool.Cytokine that is produced locally in the diseased bowel mayleak into the bowel lumen. Another possibility is that IBD-affected epithelial cells and immune-system cells, bothcapable of producing cytokines (25, 38), may be extrudedinto the bowel lumen and appear in the stool. The explanationthat the stool cytokine level is mainly responsible for theleakage from the blood vessel within the gut wall is excludedby previous data showing 100-fold higher levels of variouscytokines in stool than in plasma of patients with acuteshigellosis (16).

In conclusion, measurement of cytokines in stool mayprovide a useful and noninvasive means of understanding ofpathophysiology and clinical monitoring in IBD.

ACKNOWLEDGEMENTS

Supported in part by a grant-in-aid from the Ministry ofHealth and Welfare in Japan. The authors wish to thank themembers of the Gastroenterology Unit, Second Dept. ofMedicine, Kurume University, for their invaluable coopera-tion. Thanks are due also to Mr. Tetsuji Yamashita ofMitsubishi Kagaku Bioclinical Laboratories Inc. and to Mrs.Kiyoshi Oketani and Manabu Murakami of Eisai Co., Ltd. fortechnical assistance.

REFERENCES

1. Podolsky DK. Inflammatory bowel disease. I. N Engl J Med1991;325:928–37.

2. Fiocchi C. Cytokines. In: Targan SR, Shanahan F, editors.Inflammatory bowel disease: from bench to bedside. Baltimore:Williams & Wilkins; 1994. p. 106–22.

3. Sartor RB. Cytokines in intestinal inflammation: pathophysio-logical and clinical considerations. Gastroenterology 1994;106:533–9.

4. Reinecker H-C, Schreiber S, Stenson WF, MacDermott RP. Therole of the mucosal immune system in ulcerative colitis andCrohn’s disease. In: Ogra PL, Strober W, Mestecky J, McGheeJR, Lamm ME, Bienenstock J, editors. Handbook of mucosalimmunology: San Diego (CA): Academic Press; 1994. p. 439–56.

5. Stevens C, Walz G, Singaram C, Lipman ML, Zanker B,Muggia A, et al. Tumor necrosis factor-a, interleukin-1b, andinterleukin-6 expression in inflammatory bowel disease. Dig DisSci 1992;37:818–26.

6. Isaacs KL, Sartor RB, Haskill S. Cytokine messenger RNAprofiles in inflammatory bowel disease mucosa detected bypolymerase chain reaction amplification. Gastroenterology1992;103:1587–95.

7. Reinecker HC, Steffen M, Witthoeft T, Pflueger I, Schreiber S,MacDermott RP, et al. Enhanced secretion of tumour necrosisfactor-alpha, IL-6, and IL-1b by isolated lamina propriamononuclear cells from patients with ulcerative colitis andCrohn’s disease. Clin Exp Immunol 1993;94:174–81.

8. Niessner M, Volk BA. Altered Th1/Th2 cytokine profiles in the

intestinal mucosa of patients with inflammatory bowel disease asassessed by quantitative reversed transcribed polymerase chainreaction (RT-PCR). Clin Exp Immunol 1995;101:428–35.

9. Van Dullemen HM, Van Deventer SJH, Hommes DW, Bijl HA,Jansen J, Tytgat GNJ, et al. Treatment of Crohn’s disease withanti-tumor-necrosis factor chimeric monoclonal antibody (cA2).Gastroenterology 1995;109:129–35.

10. Schreiber S, Heinig T, Thiele HG, Raedler A. Immunoregula-tory role of interleukin 10 in patients with inflammatory boweldisease. Gastroenterology 1995;108:1434–44.

11. Braegger CP, Nicholls S, Murch SH, Stephens S, MacDonaldTT. Tumour necrosis factor alpha in stool as a marker ofintestinal inflammation. Lancet 1992;339:89–91.

12. Nicholls S, Stephens S, Braegger CP, Walker-Smith JA,Macdonald TT. Cytokines in stools of children with inflamma-tory bowel disease or infective diarrhoea. J Clin Pathol 1993;46:757–60.

13. Truelove SC, Witts LJ. Cortisone in ulcerative colitis. Finalreport on a therapeutic trial. Br Med J 1955;2:1041–8.

14. Matts SGF. The value of rectal biopsy in the diagnosis ofulcerative colitis. Q J Med 1961;120:393–407.

15. Myren J, Bouchier IAD, Watkinson G, Softley A, Clamp SE, deDombal FT. The O.M.G.E. multinational inflammatory boweldisease survey 1976–1982. A further report on 2,657 cases.Scand J Gastroenterol 1984;19 Suppl 95:1–27.

16. Raqib R, Wretlind B, Andersson J, Lindberg AA. Cytokinesecretion in acute shigellosis is correlated to disease activity anddirected more to stool than to plasma. J Infect Dis 1995;171:376–84.

17. Raqib R, Lindberg AA, Bjo¨rk L, Bardhan PK, Wretlind B,Andersson U, et al. Down-regulation of gamma interferon,tumor necrosis factor type 1, interleukin 1 (IL-1) type 1, IL-3,IL-4, and transforming growth factorb type 1 receptors at thelocal site during the acute phase of shigella infection. InfectImmunol 1995;63:3079–87.

18. Hirao J, Hibi S, Andoh T, Ichimura T. High levels of circulatinginterleukin-4 and interleukin-10 in Kawasaki disease. Int ArchAllergy Immunol 1997;112:152–6.

19. Howard M, O’Garra A, Ishida H, de-Waal-Malefyt R, de-VriesJ. Biological properties of interleukin 10. J Clin Immunol1992;12:239–47.

20. Adeyemi EO, Hodgson HJF. Faecal elastase reflects diseaseactivity in active ulcerative colitis. Scand J Gastroenterol 1992;27:139–42.

21. Mahida YR, Wu K, Jewell DP. Enhanced production ofinterleukin 1-beta by mononuclear cells isolated from mucosawith active ulcerative colitis of Crohn’s disease. Gut 1989;30:835–8.

22. Mitsuyama K, Sasaki E, Toyonaga A, Ikeda H, Tsuruta O, IrieA, et al. Colonic mucosal interleukin-6 in inflammatory boweldisease. Digestion 1991;50:104–11.

23. Mitsuyama K, Toyonaga A, Sasaki E, Watanabe K, Tateishi H,Nishiyama T, et al. IL-8 as an important chemoattractant forneutrophils in ulcerative colitis and Crohn’s disease. Clin ExpImmunol 1994;96:432–6.

24. Nishiyama T, Mitsuyama K, Toyonaga A, Sasaki E, TanikawaK. Colonic mucosal interleukin 1 receptor antagonist ininflammatory bowel diseae. Digestion 1994;55:368–73.

25. Casini-Raggi V, Kam L, Chong YJT, Fiocchi C, Pizarro TT,Cominelli F. Mucosal imbalance of IL-1 and IL-1 receptorantagonist in inflammatory bowel disease. J Immunol 1995;154:2434–40.

26. MacDonald TT, Hutchings P, Choy MY, Murch S, Cooke A.Tumour necrosis factor-alpha and interferon-gamma productionmeasured at the single cell level in normal and inflamed humanintestine. Clin Exp Immunol 1990;81:301–5.

27. Nielsen OH, Køppen T, Rudiger N, Horn T, Eriksen J, Kirman I.Involvement of interleukin-4 and -10 in inflammatory boweldisease. Dig Dis Sci 1996;41:1786–93.

28. Fuss IJ, Neurath M, Boirivant M, Klein JS, de la Motte C,Strong SA, et al. Disparate CD4� lamina propria (LP)lymphokine secretion profiles in inflammatory bowel disease.Crohn’s disease LP cells manifest increased secretion of IFN-g,

Stool Cytokine Levels in IBD 621

Scan

d J

Gas

troe

nter

ol D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsita

t Aut

onom

a B

arce

lona

on

10/2

7/14

For

pers

onal

use

onl

y.

whereas ulcerative colitis LP cells manifest increased secretionof IL-5. J Immunol 1996;157:1261–70.

29. Radford-Smith G, Jewell DP. Cytokines and inflammatorybowel disease. Baillieres Clin Gastroenterol 1996;10:151–64.

30. Dinarello CA, Thompson RC. Blocking IL-1: interleukin 1receptor antagonist in vivo and in vitro. Immunol Today 1991;12:404–10.

31. Beutler B, Cerami A. The biology of cachectin/TNF—a primarymediator of the host response. Annu Rev Immunol 1989;7:625–55.

32. Hart PH, Vitti GF, Burgess DR, Whitty GA, Piccoli DS,Hamilton JA. Potential antiinflammatory effects of interleukin-4: Suppression of human monocyte tumor necrosis factora,interleukin 1, and prostaglandin E2. Proc Natl Acad Sci USA1989;86:3803–7.

33. Vannier E, Miller LC, Dinarello CA. Coordinated antiinflam-matory effects of interleukin 4: interleukin 4 suppressesinterleukin 1 production but up-regulates gene expression and

synthesis of interleukin 1 receptor antagonist. Proc Natl AcadSci USA 1992;89:4076–80.

34. Fiorentino DF, Zlotnik A, Mosmann TR, Howard M, O’Garra A.IL-10 inhibits cytokine production by activated macrophages. JImmunol 1991;147:3815–22.

35. Howard M, O’Garra A. Biological properties of interleukin 10.Immunol Today 1992;13:198–200.

36. Zimmerman MJ, Jewell DP. Cytokines and mechanisms ofaction of glucocorticoids and aminosalicylates in the treatmentof ulcerative colitis and Crohn’s disease. Aliment PharmacolTher 1996;10:93–8.

37. Barnes PJ, Adcock I. Anti-inflammatory actions of steroids:molecular mechanisms. Trends Pharmacol Sci 1993;14:436–41.

38. Eckmann L, Jung HC, Schu¨rer-Maly C, Panja A, Morzycka-Wroblewska E, Kagnoff MF. Differential cytokine expressionby human intestinal epithelial cell lines: regulated expression ofinterleukin 8. Gastroenterology 1993;105:1689–97.

Received 16 December 1997Accepted 18 February 1998

622 T. Saiki et al.

Scan

d J

Gas

troe

nter

ol D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsita

t Aut

onom

a B

arce

lona

on

10/2

7/14

For

pers

onal

use

onl

y.