Retrograde Ileoscopy in Chronic NonbloodyDiarrhea: A Prospective, Case-Control StudySergio Morini, M.D., Roberto Lorenzetti, M.D., Francesca Stella, M.D., Maria Teresa Martini, M.D.,Cesare Hassan, M.D., and Angelo Zullo, M.D.Gastroenterology and Digestive Endoscopy, “Nuovo Regina Margherita” Hospital, Rome, Italy; andPathology, “S. Giacomo” Hospital, Rome, Italy

OBJECTIVE: Chronic nonbloody diarrhea (CND) is a fre-quent intestinal disorder, with a relevant economic impact.Besides colonic diseases, alterations of the terminal ileumcould be involved in CND pathogenesis. The aim of thisstudy was to assess the role of retrograde ileoscopy withbiopsy in CND patients.

METHODS: Patients complaining of CND and matched con-trol subjects were enrolled in the study. Retrograde ileos-copy with biopsy was attempted in all cases. Endoscopicand histological features of Crohn’s disease, nonspecificileitis, and nodular lymphoid hyperplasia were recorded foreach patient. Exclusion criteria were presence of any colonicalterations at either endoscopy or histology as well as failureof ileal intubation.

RESULTS: Overall, 156 patients were recruited. Ileal intuba-tion was successful in 149 (95.5%), but 11 (7%) patientswere excluded because colonic diseases were detected athistology. At endoscopy, alterations of the terminal ileumwere significantly more frequent in patients than in controls(47/138 vs 15/138; p � 0.0001). Crohn’s disease (9/138 vs0/138; p � 0.007) and nonpecific ileitis (18/138 vs 2/138; p� 0.0009) were significantly more frequent in patients thanin controls as well as nodular lymphoid hyperplasia (33/138vs 16/138; p � 0.008). A final diagnosis of Crohn’s diseasewas achieved on the basis of both endoscopic and histolog-ical findings in eight (5.8%) patients.

CONCLUSIONS: Retrograde ileoscopy is an useful procedurein CND because of its ability to detect alterations in theterminal ileum. Its inclusion in diagnostic workup should beconsidered. (Am J Gastroenterol 2003;98:1512–1515. ©2003 by Am. Coll. of Gastroenterology)

INTRODUCTION

Chronic nonbloody diarrhea (CND) is a fairly commonintestinal disorder, with a relevant economic impact (1).Several pathological mechanisms may be responsible forthis condition (2, 3). Once clinical examination is unreward-ing and intestinal infection has been excluded, colonoscopywith biopsy is usually performed (1). However, alterationsof the colonic mucosa have been reported to be involved in

CND pathogenesis only in 15�18% of cases (4). On theother hand, it is well known that CND may be caused bydiseases involving the terminal ileum (5). Indeed, retrogradeileoscopy was shown to be highly useful when applied tospecific clinical conditions, such as inflammatory boweldisease (6), HIV seroposivity (7), tuberculosis (8), and se-ronegative spondylarthropathy (9). However, no study hassystematically evaluated the role of ileoscopy with biopsy inCND patients without colonic involvement.

The aim of this prospective case-control study was toassess the role of endoscopic and histological examinationof the terminal ileum in CND patients.

MATERIALS AND METHODS

Subjects and ProtocolAll patients referred for colonoscopy because of CND wereconsidered for enrollment in the study. CND was defined asproduction of loose nonbloody stools with or without in-creased stool frequency and lasting more than 4 wk, assuggested in the AGA guidelines (1). For the purpose of thestudy, all patients underwent routine biochemical tests in-cluding full blood count, erythrocyte sedimentation rate,C-reactive protein, amylase, lipase, serum immunoglobulinconcentrations, antigliadin (IgA and IgG) and antiendomy-sial antibodies (IgA), antibodies to HIV, and thyroid hor-mones. Stool analysis and fecal culture was performed in allpatients, and the presence of parasites, cysts, and ova wasruled out by fecal examination. Patients in whom eitherbiochemical or fecal examinations were abnormal, thosewith an already known diarrhea-associated disorder such asinflammatory bowel disease or tuberculosis or those com-plaining of arthropathy or with previous GI surgery were notincluded. For study purposes, those patients with any endo-scopic alterations of the colonic mucosa that were poten-tially associated with CND, such as ulcerative or indeter-minate colitis, Crohn’s colitis, diverticular-associatedsigmoiditis, pseudomembranous or ischemic colitis, werealso excluded. Patients using nonsteroid anti-inflammatorydrugs were considered ineligible. In all patients, retrogradeileal intubation was attempted. All endoscopic examinationswere performed by two experienced endoscopists (S.M.,

THE AMERICAN JOURNAL OF GASTROENTEROLOGY Vol. 98, No. 7, 2003© 2003 by Am. Coll. of Gastroenterology ISSN 0002-9270/03/$30.00Published by Elsevier Inc. doi:10.1016/S0002-9270(03)00305-8

R.L.). Endoscopic examinations were performed using avideocolonoscope Olympus CF type Q160AI (Olympus,Milan, Italy). Endoscopic alterations of the terminal ileumwere carefully recorded. In detail, the following findingswere taken into account: 1) presence of features suggestiveof Crohn’s disease (aphtoid ulcer, superficial or deep ulcers,nonulcerated or ulcerated stenosis, pseudopolyps, healedulceration) (10); 2) nonspecific inflammatory alterations ofthe mucosa (minute erosions, frank erythema, mucosalswelling); and 3) nodular lymphoid hyperplasia. Using stan-dard biopsy forceps, biopsy samples were taken from thefour quadrants, including the mesenteric and antimesentericborders, and the ileum (four specimens), and additionalsamples were taken in the colon from the cecum, transversecolon, and rectum (two specimens from each site). Crohn’sdisease at histology was diagnosed according to previouslyreported criteria (10). Nonspecific ileitis was defined as thepresence of an increased lamina propria inflammatory infil-trate without features of cryptitis or granulomata. All his-tological slides were reviewed by two experienced pathol-ogists (M.T.M., F.S.) who were unaware of the clinical andendoscopic data. As control group, age- and sex-matchedpatients without CND were enrolled. Controls underwentthe same endoscopic and bioptic procedures as patients.Written informed consent was obtained for each patientbefore endoscopic procedure.

Statistical AnalysisStatistical analysis was carried out using �2 and Fisher’sexact tests, as appropriate. Differences were consideredsignificant at a 5% probability level.

RESULTS

A total of 156 consecutive patients with CND were consid-ered for enrollment in the study. Seven (4.5%) patientswere not included because of ileal intubation failure, and11 (7%) patients were excluded because histological as-sessment of the colon revealed a feature of microscopiccolitis (10 patients) or collagenous colitis (one patient).Hence, the final study population included 138 patients(82 men, 56 women; mean age 41.8 yr, range 18 –71 yr).An equal number of matched controls were included (82men, 56 women; mean age 42.9 yr, range 18 –72). Themedian duration of CND was 6 months (range 2–13months), and the median number of bowel movementswas four (range one to six). Indications for colonoscopyin the control group are listed in Table 1.

Ileoscopy revealed macroscopic alterations in 47 (34%)patients. Nine (6.5%) patients had an ileal feature suggestiveof Crohn’s disease. Histology confirmed the diagnosis ineight (5.8%) of them, and nonspecific ileitis was detected inthe remaining patient. A macroscopic feature of nonspecificinflammatory alterations was depicted in 18 (13%) patients,resulting, at histology, in nonspecific ileitis in 14 patientsand in normal mucosa in four. In the remaining 20 patients,

an endoscopic diagnosis of diffuse nodular lymphoid hy-perplasia was performed and confirmed by histology in allcases. With regard to the 91 patients with normal ilealmucosa at endoscopy, histology detected nonspecific ileitisin two (2.2%) cases, nodular lymphoid hyperplasia in 13(14.3%), and a normal finding in the remaining 76 patients.Overall, architectural alterations and irregularities of thevilli were absent in all patients diagnosed with nonspecificileitis, whereas the inflammatory infiltrate was active in allbut two cases.

In the control group, endoscopic features suggestive ofCrohn’s disease were not detected, whereas nonspecificinflammatory alterations were described in two (1.5%)cases, nodular lymphoid hyperplasia in 13 (9.4%), and nor-mal findings in the remaining 123 subjects. Histology de-tected nonspecific ileitis in 5 (3.6%) subjects, two in thosewith endoscopic alterations and three in those without it, andnodular lymphoid hyperplasia in 16 (11.6%) cases.

The comparison between the two groups showed that theoverall endoscopic alterations were significantly morefrequent in patients than in controls (47/138 vs 15/138; p� 0.0001). The prevalence of both Crohn’s disease (9/138 vs 0/138; p � 0.007) and nonspecific inflammatoryalterations (18/138 vs 2/138; p � 0.0009) was signifi-cantly higher in patients than in controls, whereas nodularlymphoid hyperplasia was similarly distributed in the twogroups (20/138 vs 13/138; p � 0.2). Likewise, the overallhistological lesions were detected more frequently inpatients than in controls (58/138 vs 21/138; p � 0.0001).In particular, both Crohn’s disease (8/138 vs 0/138; p �0.01) and nonspecific ileitis (17/138 vs 5/138; p � 0.01)occurred more frequently in patients than in controls, aswell as nodular lymphoid hyperplasia (33/138 vs 16/138;p � 0.008). Subjects harboring lymphoid hyperplasia (n� 49) were significantly younger than those without it(36.7 vs 44.9; p � 0.0001). The mean age differenceemerged in both patients (35. 8 vs 43.7; p � 0.0001) andcontrols (38.6 vs 43.5; p � 0.069). A final diagnosis ofCrohn’s disease was made on the basis of both endo-scopic and histological findings in eight (5.8%) patients.All of them underwent small bowel follow-through andbowel wall ultrasound, which showed alterations sugges-tive of Crohn’s disease in three cases and normal findingsin the remaining five.

Table 1. Indications for Endoscopy in Control Subjects WithoutChronic Nonbloody Diarrhea.

Indication Number of Cases

Abdominal pain 35Adenoma follow-up 53Colorectal screening 18Rectal bleeding 20Constipation 12Total 138

1513AJG – July, 2003 Ileoscopy in Chronic Diarrhea

DISCUSSION

Retrograde ileoscopy has become a relatively easy, painless,and rapid procedure that is feasible in most cases. Recentdata indicate a high intubation rate of the terminal ileumranging from 79% to 95%, adding no more than 3–4 min tothe entire duration of colonoscopy without increasing therisk of complications (7, 11�13). In clinical setting, ileos-copy has been regarded as a meaningful procedure in spe-cific conditions, including inflammatory and some infec-tious diseases (6–9). Indeed, the availability of thehistological examination of the ileal mucosa may allow anaccurate and definitive diagnosis. All these observationscharacterize ileoscopy as an appealing diagnostic tool incomparison to the widely applied radiological techniquesfor the assessment of the terminal ileum.

CND is a frequent bowel disorder in patients presenting togeneral practitioners or referred to gastroenterologists and isexperienced by 3–5% of the general population (14–16).Although it is difficult to estimate the economic impact ofCND, the overall cost of work loss, disability payments, andlost productivity has been suggested to be remarkably high(1). Current guidelines advise performing a variety of testsbefore classifying these patients as having irritable bowelsyndrome, inasmuch as defined causes may be detected andeffectively treated (17). After negative findings at both clin-ical examination and biochemistry, colonoscopy with bi-opsy is advised (5). However, colonoscopy can identify aspecific etiology of CND in 30% of the cases at best (18).The role of the ileoscopy in defining the pathogenesis ofchronic diarrhea has been evaluated in a few studies. More-over, these studies included patients with symptoms sug-gestive of inflammatory bowel disease (6) or those withunselected indications (7, 8), and in some patients HIVseropositivity or tuberculosis infection was present. Weaimed to evaluate whether retrograde ileoscopy could rep-resent a significant diagnostic gain in selected patients withCND without other symptoms or signs suggestive of anyspecific disease.

In our study, systematic ileal intubation led to detection ofalterations of the terminal ileum in more than one third ofpatients. Interestingly, a final diagnosis of Crohn’s diseasewas achieved in 5.8% of patients, some of whom wouldhave been otherwise dismissed with a diagnosis of irritablebowel disease. Indeed, subsequent radiological examinationfailed to detect alterations in five of the eight cases in whomCrohn’s disease was identified. Therefore, ileoscopy sub-stantially affected the management of a subgroup of CNDpatients deserving specific treatment and follow-up. Intrigu-ingly, even the prevalence of nonspecific inflammatorychanges of the ileal mucosa was significantly higher inpatients than in controls, both at endoscopy and at histology,indicating a possible role in CND pathogenesis. Are thesepatients developing inflammatory bowel diseases? Is this thesign of an indefinite infectious disease? The clinical mean-ing of these nonspecific alterations, as well as possible

therapeutic approaches, needs further investigation. All ofthese patients are in ongoing follow-up.

Although nodular lymphoid hyperplasia occured in nearly10% of the control subjects, it was encountered much morefrequently in the patient group. Does nodular lymphoidhyperplasia have a role in CND pathogenesis? Suggestively,this finding has been claimed to have a role in chronicdiarrhea associated with some specific clinical conditions,such as common variable hypogammaglobulinemia (19, 20)and in children with developmental disorders (21). Intrigu-ingly, subjects harboring nodular lymphoid hyperplasiawere significantly younger than those without it. On thecontrary, our data showed that ileoscopy routinely per-formed in subjects without diarrhea is clinically meaning-less, as reported in few previous studies (7, 11).

We therefore conclude that retrograde ileoscopy has anappealing role in the diagnosis of CND, expanding thepotential indications for this procedure. This proposal isfurther supported by its technical feasibility, which results ina safe and painless procedure that is successful in more than95% of cases. Moreover, ileoscopy with biopsy has beenproved, in our study, to be more accurate than radiology indetecting Crohn’s disease in an early stage, as reportedelsewhere (22, 23). It can be speculated that these endo-scopic alterations of the terminal ileum can be detected bythe innovative wireless endoscopy or by the new imagingprocedures (24) but that the final diagnosis would alwaysrequire histological confirmation. Therefore, these observa-tions suggest that ileo-colonoscopy, instead of simplecolonoscopy, should be included in the flow-chart for CND,clearly improving the effectiveness of the diagnosticworkup.

In conclusion, this is the first large, prospective, case-control study showing the usefulness of retrograde ileos-copy in CND because of the ability of this technique todetect organic diseases in a subgroup of patients requiringspecific management.

Reprint requests and correspondence: Sergio Morini, M.D.,Ospedale Nuovo Regina Margherita, Gastroenterologia ed En-doscopia Digestiva, Via Morosini 30, 00153, Roma, Italia.

Received July 22, 2002; accepted Jan. 14, 2003.

REFERENCES

1. Fine KD, Schiller LR. AGA technical review on the evaluationand management of chronic diarrhea. Gastroenterology 1999;116:1464–86.

2. Read NW, Krejs GJ, Read MG, et al. Chronic diarrhea ofunknown origin. Gastroenterology 1980;78:264–71.

3. Bytzer P, Stokholm M, Andersen I, et al. Aetiology, medicalhistory and fecal weight in adult patients referred for diar-rhoea: A prospective study. Scand J Gastroenterol 1990;25:572–8.

4. Patel Y, Pettigrew NM, Grahame GR, Bernstein CN. Thediagnostic yield of lower endoscopy plus biopsy in nonbloodydiarrhea. Gastrointest Endosc 1997;46:338–43.

5. Fine KD, Krejs GJ, Fordtran JS. Diarrhea. In: Sleisenger MH

1514 Morini et al. AJG – Vol. 98, No. 7, 2003

and Fordtran JF, eds. Gastrointestinal disease. Philadelphia:WB Saunders, 1993:1043–72.

6. Geboes K, Ectors N, D’Haens G, Rutgeerts P. Is ileoscopywith biopsy worthwhile in patients presenting with symptomsof inflammatory bowel disease. Am J Gastroenterol 1998;93:201–6.

7. Zwas FR, Bonheim NA, Berken CA, Gray S. Diagnostic yieldof routine ileoscopy. Am J Gastroenterol 1995;90:1441–3.

8. Bhasin DK, Goenka MK, Dhavan S, et al. Diagnostic value ofileoscopy. A report from India. J Clin Gastroenterol 2000;31:144–6.

9. Bedert LC, De Clerk LS, Stevens DJ, et al. Ileocolonoscopy inpatients with spondylarthritis and/or asymmetric oligoarthritis.Gastroenterology 1990;100:1005 (letter).

10. Farmer M, Petras RE, Hunt LE, et al. The importance ofdiagnostic accuracy in colonic inflammatory bowel disease.Am J Gastroenterol 2000;95:3184–8.

11. Kundrotas LW, Clement DJ, Kubik CM, et al. A prospectiveevaluation of successful terminal ileum intubation during rou-tine colonoscopy. Gastrointest Endosc 1994;40:544–6.

12. Marshall JB, Barthel JS. The frequency of total colonoscopyand terminal ileal intubation in the 1990s. Gastrointest Endosc1993;39:518–20.

13. Iacopini G, Vitale MA, De Cesare MA, et al. Routine retro-grade ileoscopy: A prospective evaluation of successful ter-minal ileum intubation during conventional colonoscopy. Di-gest Liver Dis 2002;34(suppl 1):A112 (abstract).

14. Talley NJ, O’Keefe EA, Zinsmeister AR, Melton LJ III. Prev-alence of gastrointestinal symptoms in the elderly: A popula-tion-based study. Gastroenterology 1992;102:895–901.

15. Talley NJ, Weaver AL, Zinsmeister AR, Melton LJ III. Onset

and disappearance of gastrointestinal symptoms and functionalgastrointestinal disorder. Am J Epidemiol 1992;136:165–77.

16. Lembo T, Wright RA, Bagby B, et al. Alosetron controlsbowel urgency and provides global symptom improvement inwomen with diarrhea predominant irritable bowel syndrome.Am J Gastroenterol 2001;96:2662–70.

17. American Gastroenterological Association. American Gastro-enterological Association Medical Position Statement. Guide-lines for the evaluation and management of chronic diarrhea.Gastroenterology 1999;116:1461–4.

18. Shah RJ, Fenoglio-Preiser C, Bleau BL, Giannella RA. Use-fulness of colonoscopy with biopsy in the evaluation of pa-tients with chronic diarrhea. Am J Gastroenterol 2001;96:1091–5.

19. Hermaszewsky RA, Webster ADB. Primary hypogamma-globulinemia: A survey of clinical manifestations and compli-cations. Q J Med 1993;86:31–42.

20. Hermans PE, Diaz-Buxo, Stobo JD. Idiopathic late-onset im-munoglobulin deficiency. Am J Med 1976;61:221–36.

21. Wakefield AJ, Anthony A, Murch SH, et al. Enterocolitis inchildren with developmental disorders. Am J Gastroenterol2000;95:2285–95.

22. Tribl B, Turetschek K, Mostbeck G, et al. Conflicting resultsof ileoscopy and small bowel double-contrast barium exami-nation in patients with Crohn’s disease. Endoscopy 1998;30:339–44.

23. Lewis BS. Ileoscopy should be part of standard colonoscopy.J Clin Gastroenterol 2000;31:103–4.

24. Hassan C, Cerro P, Zullo A, et al. Computed tomography (CT)enteroclysis in comparison to ileo-colonoscopy in patientswith Crohn’s disease (CD). Int J Colorectal Dis 2003;18:121–5.

1515AJG – July, 2003 Ileoscopy in Chronic Diarrhea