Search

Objective Some patients (10 % – 32 %) with a positive guaiac fecal occult blood test (gFOBT) do not undergo the recommended colonoscopy. The aim of this study was to compare video capsule endoscopy (VCE) and computed tomography colonography (CTC) in terms of participation rate and detection outcomes when offered to patients with a positive gFOBT who did not undergo the recommended colonoscopy. Methods An invitation letter offering CTC or VCE was sent to selected patients after randomization. Acceptance of the proposed (or alternative) procedure and procedure results were recorded. Sample size was evaluated according to the hypothesis of a 13 % increase of participation with VCE. Results A total of 756 patients were targeted. Following the invitation letter, 5.0 % (19/378) of patients underwent the proposed VCE and 7.4 % (28/378) underwent CTC, ( P = 0.18). Following the letter, 9.8 % (37/378) of patients in the VCE group underwent a diagnostic procedure (19 VCE, 1 CTC, 17 colonoscopy) vs. 10.8 % in the CTC group (41/378: 28 CTC, 13 colonoscopy; P = 0.55). There were more potentially neoplastic lesions diagnosed in the VCE group than in the CTC group (12/20 [60.0 %] vs. 8/28 [28.6 %]; P = 0.04). Thus, 15/20 noninvasive procedures in the VCE group (19 VCE, 1 CTC; 75.0 %) vs. 10/28 in the CTC group (35.7 %; P = 0.01) resulted in a recommendation of further colonoscopy, but only 10/25 patients actually underwent this proposed colonoscopy. Conclusion Patients with a positive gFOBT result who do not undergo the recommended colonoscopy are difficult to recruit to the screening program and simply proposing an additional, less-invasive procedure, such as VCE or CTC, is not an effective strategy. ClinicalTrials.govNCT02558881TRIAL REGISTRATION: Randomized, controlled trial NCT02558881 at clinicaltrials.gov.

Esophagogastroduodenoscopy (EGD) is the standard method for the diagnosis of esophago-gastric varices. The aim of this prospective multicenter study was to evaluate the PillCam esophageal capsule endoscopy (ECE) for this indication.Patients presenting with cirrhotic or noncirrhotic portal hypertension underwent ECEfollowed by EGD at the time of diagnosis. Capsule recordings were blindly read by two endoscopists.A total of 120 patients (72 males, mean age: 58 years; mean Child-Pugh score: 7.2) were included. Esophageal varices were detected in 74 patients. No adverse event was observed after either EGD or ECE. Seven (6%) patients were unable to swallow the capsule. The mean recording time was 204 s (range 1-876). Sensitivity, specificity, negative predictive value, and positive predictive value of ECE for the detection of esophageal varices were 77%, 86%, 69%, and 90%, respectively. Sensitivity, specificity, negative and positive predictive values of ECE for the indication of primary prophylaxis (esophageal varicesor = grade 2 and/or red signs) were 77, 88, 90, and 75%, respectively, and 85% of the patients were adequately classified for the indication (or not) of prophylaxis. Interobserver concordance for ECE readings was 79.4% for the diagnosis of varices, 66.4% for the grading of varices, and 89.7% for the indication of prophylaxis.This large multicenter study confirms the safety and acceptable accuracy of ECE for the evaluation of esophageal varices. ECE might be proposed as an alternative to EGD for the screening of portal hypertension, especially in patients unable or unwilling to undergo EGD.

BACKGROUND AND STUDY AIM: Polyp miss rates during colonoscopy have been calculated in a few tandem or back-to-back colonoscopy studies. Our objective was to assess the adenoma miss rate while limiting technique or operator expertise biases, i. e. by performing a large multicenter study, with same-day back-to-back video colonoscopy, done by two different operators in randomized order and blinded to the other examination. PATIENTS AND METHODS: 294 patients at 11 centers were included. Among the 286 analyzable tandem colonoscopies, miss rates were calculated in both a lesion- and patient-based analysis. Each of these rates was determined for polyps overall, for adenomas, and then for lesions larger than 5 mm, and for advanced adenomas. Univariate and logistic regression analysis were performed to define independent variables associated with missed polyps or adenomas. RESULTS: The miss rates for polyps, adenomas, polyps ≥ 5 mm, adenomas ≥ 5 mm, and advanced adenomas were, respectively, 28, 20, 1, 9 and 11 %. None of the masses with a carcinomatous (n = 3) or carcinoid component (n = 1) was missed. The specific lesion miss rates for patients with polyps and adenomas were respectively 36 and 26 % but the corresponding rates were 23 and 9.4 % when calculated for all 286 patients. The diameter (1-mm increments) and number of polyps (≥ 3) were independently associated with a lower polyp miss rate, whereas sessile or flat shape and left location were significantly associated with a higher miss rate. Adequacy of cleansing, presence of diverticula, and duration of withdrawal for the first procedure were not associated with adenoma miss rate. CONCLUSIONS: We confirm a significant miss rate for polyps or adenoma during colonoscopy. Detection of flat polyps is an issue that must be focused on to improve the quality of colonoscopy.

BACKGROUND AND STUDY AIMS: The high cumulative risk of colorectal cancer in patients with familial adenomatous polyposis (FAP) justifies prophylactic colectomy with either ileorectal (IRA) or ileal-pouch-anal anastomosis (IPAA). Our aim was to evaluate retrospectively the frequency of and time interval to adenoma development in the ileal mucosa of patients with both types of surgery. PATIENTS AND METHODS: Retrospective study of 44 FAP patients with IRA (n = 21) and IPAA (n = 23). All patients were followed with a standardized procedure including chromoscopy and biopsies of visible polyps. In patients with IRA, specific attention was paid to the ileal mucosa above the anastomosis. RESULTS: In the IPAA group, 18/23 patients (78 %) presented with visible polyps [histology: 16 (70 %) had adenoma with low-grade dysplasia; 1 (4 %) had adenoma with high-grade dysplasia; 1 had normal mucosa]. The mean interval between colectomy and the diagnosis of adenoma was 4.7 +/- 3.3 years. In the IRA group, 16/21 patients (77 %) presented visible polyps in the ileal mucosa [adenoma with low-grade dysplasia in 8 patients (38 %), with high-grade dysplasia in 2 (10 %), and lymphoid nodular hyperplasia in 6]. The mean interval between colectomy and adenoma diagnosis was significantly shorter in the IPAA than in the IRA group (4.76 +/- 3.3 vs. 16.4 +/- 8.5 years, P< 0.0001). CONCLUSION: Our results show a high frequency of adenomas in the ileal mucosa of patients with IPAA and IRA (74 % and 48 % respectively), with evolution into high-grade dysplasia in 6.7 % of cases.

BACKGROUND AND STUDY AIMS Colonoscopy is still considered the standard investigation for the detection of colorectal adenomas, but the miss rate, especially for small and flat lesions, remains unacceptably high. Chromoscopy has been shown to increase the yield for lesion detection in inflammatory bowel disease. The aim of this randomized prospective study was to determine whether a combination of chromoscopy and structure enhancement could increase the adenoma detection rate in high-risk patients. PATIENTS AND METHODS All patients included in the trial had a personal history of colorectal adenomas and/or a family history of colorectal cancer (but excluding genetic syndromes). They were randomized to one of two tandem colonoscopy groups, with the first pass consisting of conventional colonoscopy for both groups, followed by either chromoscopy and structure enhancement (the "study" group) or a second conventional colonoscopy (the control group) for the second-pass colonoscopy. All detected lesions was examined histopathologically after endoscopic resection or biopsy. The principal outcome parameter was the adenoma detection rate; the number, histopathology, and location of lesions was also recorded. RESULTS A total of 292 patients were included in the study (146 patients in each group). The patients' demographic characteristics, the indications for colonoscopy, and the quality of bowel preparation were similar in the two groups. There was a significant difference between the two groups with respect to the median duration of the examination (18.9 minutes in the control group vs. 27.1 minutes for the study group, P < 0.001). Although more hyperplastic lesions were detected throughout the colon in the study group ( P = 0.033), there was no difference between the two groups in either the proportion of patients with at least one adenoma or in the total number of adenomas detected. Chromoscopy and structure enhancement diagnosed significantly more diminutive adenomas (< 5mm) in the right colon, compared with controls ( P = 0.039). CONCLUSIONS On the basis of our results we cannot generally recommend the systematic use of chromoscopy and structure enhancement in a high-risk patient population, although the detection of small adenomas in the proximal colon was improved.

Presentee pour la premiere fois a San Francisco en mai 2000, la capsule endoscopique permet depuis 2001 en Europe, l’exploration non invasive et complete de la totalite de l’intestin grele. Sa distribution par la societe Given Imaging LTD a ete rapide pour atteindre 50 000 capsules ingerees en Europe (200 000 dans le monde) reparties sur 758 stations de travail (2250 dans le monde). La premiere indication de videocapsule reste l’hemorragie digestive inexpliquee. Cette nouvelle technique d’exploration de l’intestin grele a su prendre ici une place preponderante par son caractere non invasif, son innocuite et sa simplicite.

The cystic presentation of endocrine tumours is rare and raises difficult diagnostic problems. So far, the only cases of cystic digestive endocrine tumours reported in the literature are of pancreatic origin. We report the unusual observation of a jejunal endocrine carcinoma presenting as a cystic abdominal mass. A 59-year-old woman was referred for chest and abdominal pain. Imaging studies revealed multiple cystic nodules in the liver and a large sus-mesocolic cystic lesion of probable intestinal origin. Biopsies of the extra-hepatic mass and liver nodules showed endocrine tumour. Surgical resection of the jejunal mass and of liver segment III were performed. Histological examination confirmed the diagnosis of jejunal endocrine carcinoma metastatic to the liver. Large areas of the primary and secondary tumours presented an unusual vesicular architecture, responsible for the cystic presentation. No adjuvant treatment was attempted. This observation underlines the difficult diagnostic problems raised by the cystic presentation of digestive endocrine tumours.

Resume Objectif. – Rechercher des facteurs de risque clinique de l'enterocele et evaluer son association avec d'autres troubles de la statique pelvienne. Patientes et methodes. – Cinq cent quarante-quatre patientes consecutives ont ete explorees par colpo-cysto-defecographie avec opacification de l'intestin grele. Cent trente-six cas de femmes avec enterocele ont ete compares a 408 cas de femmes sans enterocele. Resultats. – Il n'y avait pas de difference significative entre les deux groupes pour l'âge, les antecedents obstetricaux (parite, macrosomie fœtale, antecedents de forceps ou de dechirure lors d'un accouchement). Les femmes avec enterocele presentaient plus souvent des antecedents d'hysterectomie (60 vs 24 %) ou de cystopexie (29 vs 13 %). D'autres troubles de la statique pelvienne etaient associes dans 91 % des cas d'enterocele : rectocele (25 %), cystocele (42 %), hysterocele (28 %), prolapsus rectal interne (52 %) ou externe (4 %) et descente perineale (30 %). L'incontinence urinaire etait plus frequente chez les femmes sans enterocele. Discussion et conclusions. – Cette etude demontre que l'hysterectomie et la cystopexie favorisent la formation d'enterocele et confirme que l'enterocele est souvent associee a d'autres troubles de la statique pelvienne. L'enterocele pourrait avoir un role pessaire sur les troubles urinaires. Les resultats de la colpo-cysto-defecographie renforcent la notion de pathologie regionale du plancher pelvien.

In Lynch syndrome, flat and diminutive adenomas are particularly prone to malignant transformation, but they can be missed by standard colonoscopy. It is not known whether chromocolonoscopy is able to detect more adenomas than standard colonoscopy in patients with Lynch syndrome.We conducted a prospective, multicenter, randomized trial to compare standard colonoscopy with standard colonoscopy followed by pancolonic chromoscopy with indigo carmine in patients with a proven germline mutation in a mismatch-repair gene related to Lynch syndrome and who were undergoing screening or surveillance colonoscopy. Standard colonoscopy was used first to detect visible lesions. Colonoscopy with chromoscopy was then performed by a second gastroenterologist (blinded to the findings of the first colonoscopy) to detect additional lesions. The primary end point was the number of patients in whom at least one adenoma was detected.A total of 78 eligible patients (median age, 45 years) were enrolled at 10 centers from July 2008 to August 2009. Significantly more patients with at least one adenoma were identified by chromocolonoscopy (32/78 (41%)) than by standard colonoscopy (18/78 (23%); P0.001). The percentage of patients in whom at least one additional adenoma was detected during the chromoscopy was 31% (24/78). Overall, chromocolonoscopy plus colonoscopy detected a total of 55 adenomas in 32 patients (mean number of adenomas detected per patient: 0.7 vs. standard colonoscopy alone: 0.3; P0.001).The results support the proposition that chromocolonoscopy may significantly improve the detection rate of colorectal adenomas in patients undergoing screening or surveillance colonoscopy for Lynch syndrome.

BACKGROUND AND STUDY AIM: A video capsule similar to that used in small-bowel capsule endoscopy is now available for esophageal exploration. The aim of our study was to compare the accuracy of upper endoscopy (esophageal gastroduodenoscopy [EGE]) with esophageal capsule endoscopy (ECE) in patients at risk of esophageal squamous cell cancer (SCC). PATIENTS AND METHODS: 68 patients at risk of SCC secondary to a history of head and neck neoplasia were included in this comparison of techniques for detecting SCC and dysplasia. ECE was done using the first generation Pillcam ESO and EGE was performed in accordance with the usual practice of each center, followed by examination with 2 % Lugol staining and biopsy of unstained areas (39 neoplasia comprising 5 low grade dysplasia, 8 high grade dysplasia and 26 SCC). RESULTS: Compared with EGE with and without Lugol staining, the sensitivities of ECE for neoplasia diagnosis were 46 % and 54 %, respectively. On a per-patient basis, the sensitivity, specificity, and positive and negative predictive value of ECE were 63 %, 86 %, 77 % and 76 %, respectively, compared with EGE without staining, and 61 %, 86 %, 77 % and 73 % compared with EGE with iodine staining. Neither the ECE transit time nor the distance between the esopharyngeal line and the neoplastic lesion differed between the 21 false-negative and 18 true-positive cases diagnosed by ECE; the only difference was a smaller median diameter among false negatives ( P < 0.001). CONCLUSION: In a cohort at high risk for esophageal SCC, ECE is not sensitive enough to diagnose neoplastic lesions.

Background Patients with hereditary hemorrhagic telangiectasia (HHT) often present with recurrent anemia because of epistaxis or GI bleeding in relation to telangiectases mostly located in the stomach or small bowel. Capsule endoscopy is considered a major diagnostic tool for small-bowel diseases, but the impact of capsule endoscopy imaging on patient management in HHT is poorly understood. Objective To clarify the contribution of capsule endoscopy in selected patients with HHT. Design Prospective, descriptive study. Setting Multicenter, two university hospital tertiary-care centers, from January 2003 to June 2007. Patients This study involved 30 patients with HHT and severe anemia (hemoglobin Intervention Capsule endoscopy investigation. Main Outcome Measurements Clinical characteristics and capsule endoscopy results and their clinical consequences. Results Capsule endoscopy detected gastric and small-bowel telangiectases in 14 (46.7%) and 26 (86.7%) cases, respectively. Active bleeding was present in 36.7% of cases. Diffuse telangiectases were detected in 42.3% without correlation with age, sex, or type of HHT mutation. Further investigations were carried out as a consequence of the capsule endoscopy results in 67% of cases. Treatment, consisting mostly of endoscopic argon plasma coagulation, was scheduled in 46.7% of patients. Limitations Our population was essentially composed of patients with the ALK1 mutation. Conclusion This study shows that there is a high diagnostic yield for capsule endoscopy in selected patients with HHT. Capsule endoscopy makes possible precise mapping of lesions and has a considerable impact on the management of these selected patients by using a predefined algorithm: a limited number of accessible lesions is suitable for endoscopic treatment, whereas innumerable diffuse lesions require a medical approach. We suggest that capsule endoscopy could be a first-line, noninvasive, digestive tract examination in selected patients with HHT.

Introduction Les tumeurs carcinoides sont rares (20 a 40% des tumeurs de l'intestin grele). Elles sont souvent de petite taille (inferieures a 1cm), de localisation intra parietale a predominance ileale et seraient multifocales dans 30% des cas. Le pronostic vital des patients et leur survie peuvent dependre de l'evolution tumorale locoregionale rendant importante l'identification de la tumeur primitive avant de discuter une prise en charge chirurgicale. La videocapsule endoscopique (VCE) est un examen non invasif permettant d'explorer la totalite de l'intestin grele. Sa faisabilite et ses resultats dans la detection des tumeurs carcinoides sont en cours d'evaluation. Patients et Methodes Notre travail a porte de facon retrospective sur une serie de 9 patients (6 femmes et 3 hommes; moyenne d'âge 69,7 ans, extremes 49–73), chez qui le diagnostic de tumeur endocrine (TE) bien differencie a ete porte avec une forte presomption pour une atteinte tumorale intestinale (metastases hepatiques, peritoneales ou ganglionnaires, mesenterite retractile au scanner ou decouverte fortuite au cours d'une ileocoloscopie). Les films endoscopiques de videocapsule ont tous ete relus en aveugle par un 2eme lecteur experimente apres une lecture preoperatoire initiale par un premier lecteur. L'identification de la tumeur primitive (selon un score de probabilite faible F0, moyen F1 ou fort F2), sa description et le retentissement local du processus tumoral ont ete apprecies. Ses resultats ont ete compares a ceux de la chirurgie realisee chez tous les patients. Resultats Aucune complication n'a ete observee. La VCE a ete evacuee dans tous les cas. Le bas fond caecal n'a pas ete atteint dans 1 cas. La VCE a permis de detecter 7 tumeurs primitives (4lesions nodulaires, 2 aspects de bombement sous-muqueux et 1 ulceration) dont la taille moyenne etait de 13mm (extremes 4–20mm) sur le compte rendu anatomopathologique apres la chirurgie. Dans un cas, la tumeur primitive n'a pas ete identifiee par la VCE (taille 20mm). Dans 2 cas, des anomalies de la muqueuse ileale ont ete interpretes comme un processus ischemique sur mesenterite retractile confirme par la chirurgie. Aucune tumeur multifocale n'a ete observee. Le champ diagnostique de la VCE pour l'identification de la tumeur primitive (lesions classees F1 et F2) etait de 77%, sa sensibilite de 87,5% et sa VVP de 54%. Conclusion En cas de suspicion de TE intestinale, un examen par VCE semble faisable et pourrait permettre d'identifier la tumeur primitive. Sa presentation endoscopique est variable en fonction de son mode de developpement intra parietal et de sa taille. Sa place reste a valider de facon prospective chez un plus grand nombre de patients par rapport aux moyens d'exploration de reference que sont l'enteroscanner et Octreoscan®.

UNLABELLED BACK AND STUDY AIMS: Endoscopic mucosal resection (EMR) is used to treat premalignant and malignant digestive tract lesions. This report presents the efficacy and safety of EMR for squamous superficial neoplastic esophageal lesions. PATIENTS AND METHODS A retrospective cohort study presented data from 51 patients with 54 lesions over an 8-year period, between November 1997 and September 2005. Dysplasas or mucosal (m) T1 carcinomas were treated with repeated EMR until there was a complete local remission. Patients with submucosal (sm) T1 carcinomas were treated with repeated EMR until there was a complete local remission. Patients with submucosal (sm) T1 carcinomas or more advanced stage were offered surgery or chemoradiotherapy. RESULTS There was no mortality, perforation, or major hemorrhage, and there were three easily dilated stenoses. Of the patients, 16 had lesions graded as T1sm or more advanced and one patient was found to have normal tissue post EMR. Complete local remission was achieved in 31 of the 34 patients with dysplasia or T1 m cancers (91%). There was no distant relapse and there was local disease recurrence in eight of the 31 patients (26%). The 5-year survival rate was 95%. CONCLUSIONS EMR for squamous superficial neoplastic lesions of the esophagus is safe and provides satisfactory survival results.

Introduction La lecture d’une capsule endoscopique du grele est de 30 minutes en moyenne. Reduire ce temps de lecture, et eventuellement optimiser la detection de lesions cliniquement importantes de l’intestin grele est un objectif que voudraient atteindre plusieurs systemes de detection informatique d’images anormales, actuellement non valides, dont la fonction Quick View (QV) de la firme Given Imaging. Le but de ce travail etait d’evaluer la sensibilite et la specificite d’une lecture en mode QV, et la fiabilite de l’algorithme. Patients et Methodes 106 films cas de capsule endoscopique etaient choisis de facon aleatoire parmi les 10 derniers cas de 12 centres ayant une experience de plus de 200 cas. Une lecture commune etait organisee mobilisant 12 lecteurs experimentes de la commission capsule de la SFED, notant en mode QV seul chaque image selon un grade de pertinence (P2 : responsabilite certaine, P1 : incertaine) et le temps de la lecture. Le resultat obtenu etait compare a la lecture (LN) du meme film en mode normal, et chaque discordance entre les 2 lectures etait revue et classee par 3 experts, image par image. Resultats Le temps moyen de lecture en mode QV etait de 11,6 min (2-27). Le temps de transit moyen du grele evalue en mode QV et LN etait de 258 et 260 min, significativement correle (p eme lecture, il s’agissait de 21 faux positifs (LN 12, QV 9), de 7 lesions non vues en LN, 7 non vues en mode QV, incluant 5 doublons. La sensibilite et la specificite par patient sont indiquees en table 1. Sur 7 lesions manquees par la lecture QV, 4 seulement (chez 3 patients) etaient absentes du film QV, representant a une sensibilite theorique par patient (fiabilite de l’algorithme) de 93,5 %. Conclusion Dans cette etude, l’algorithme d’analyse d’image Quick View detectait pres de 94 % des lesions significatives, et la lecture en mode QV etait aussi sensible et specifique, en conditions de lecture reelle, qu’une lecture normale, pour un temps de lecture beaucoup plus court. D’autres etudes sont necessaires pour confirmer ce resultat, ce qui pourrait permettre a terme de proposer en premiere intention une lecture des films de capsule du grele en mode Quick View.

Introduction L’adenocarcinome duodenal est la deuxieme cause de mortalite apres colectomie dans la polypose adenomateuse familiale. Les adenomes duodenaux sont presents chez quasiment tous les patients, mais avec une severite variable classee en 4 stades (score de Spiegelman) prenant en compte le nombre, la taille, l’aspect tubulo-villeux et le degre de dysplasie. Au stade IV, l’attitude de reference est chirurgicale mais un traitement endoscopique est parfois propose. But Evaluer l’efficacite et les risques au long cours d’une attitude therapeutique endoscopique premiere pour les patients porteurs d’une polypose duodenale avancee (stade IV). Patients et Methodes Etude retrospective bicentrique d’une cohorte de patients suivis et traites de facon standardisee pour une polypose duodenale classee initialement au stade IV de Spiegelman. Population : 36 patients (16 hommes, 20 femmes) d’âge moyen de 48 ± 12 ans (extremes 22-73 ans) porteurs d’une polypose duodenale de stade IV. Le score de Spiegelman etait evalue a chaque endoscopie. La duree moyenne de la maladie etait de 22 ± 10 ans (3-43) et celle du suivi endoscopique etait de 11 ± 6 ans (0,5-20). Resultats Le nombre moyen d’endoscopies par patient etait de 11,8 ± 6,6 (2 - 26).). Le delai moyen de l’apparition du stade IV par rapport au debut du suivi etait de 4,7 ± 4,8 (mediane de 4, extremes 0 - 17). Apres traitement endoscopique du stade IV, le recul moyen du suivi etait de 6 ± 3,8 ans (0,5 - 17, mediane 6). Le score moyen de Spiegelman initial etait de 9,9 ± 1 points (9 - 12) au moment du diagnostic du stade IV et de 5,8 ± 1,7 (0 - 8) apres traitement (p = 0,002). Les traitements endoscopiques employes etaient : laser YAG (2 patients), phototherapie dynamique (6 patients), argon (27 patients), argon et mucosectomie (7 patients) mucosectomie (18 patients). Douze patients presentaient un ampullome en DBG, 13 en DHG et 1 patient avec 1 foyer d’adenocarcinome. Traitement des neoplasies ampullaires : laser (2 patients), sphincterotomie (4 patients), ampullectomie (15 patients). 6 % (12/197) des endoscopies therapeutiques etaient suivies d’une complication : hemorragie retardee (7), pancreatite (4) et perforation duodenale (1) ayant necessite un drainage chirurgical. Trois patients ont du etre operes (1 ampullectomie et 2 resections duodeno-jejunales). Un patient traite est en attente de controle. Au cours du suivi, 4 patients sont decedes de : cholangiocarcinome (2 pts), cancer du rectum metastatique (1 pt), cancer pulmonaire (1 pt). Aucun adenocarcinome duodenal invasif n’a ete diagnostique. Conclusion Avec un recul moyen de 6 ans, la prise en charge endoscopique des patients porteurs d’une polypose duodenale avancee (stade IV) apparait etre une alternative a la chirurgie, aboutissant a une diminution du stade chez 95 % des patients, sans aucun adenocarcinome invasif diagnostique.

Introduction En France, la strategie de depistage des polypes et du cancer colorectal (CCR) repose sur la combinaison de la recherche de sang dans les selles et de la coloscopie. La video-capsule specifique du colon (VCC), pourrait etre une alternative a cette strategie [1, 2]. Le but de cette etude prospective, multicentrique francaise etait de comparer les performances diagnostiques de la VCC a celles de la coloscopie dans une population a risque moyen ou eleve de CCR. (analyse intermediaire planifiee apres 105 inclusions/550 prevues). Patients et Methodes Les patients a risque moyen de CCR (50-74 ans) ou a risque eleve (antecedents personnels ou familiaux de neoplasie colique) ont ete inclus dans l’etude. Les criteres d’exclusion etaient une symptomatologie evocatrice de stenose intestinale, une diverticulose colique recemment compliquee et une insuffisance cardiaque et/ou renale evoluee. La VCC et la coloscopie ont ete realisees, chez le meme patient, a 24 heures d’intervalle, selon une methodologie publiee (1, 2). Les donnees ont ete interpretees (niveau de preparation, presence de lesion,…), en double insu, par l’endoscopiste ayant realise la coloscopie et le lecteur de la VCC. En cas de discordance entre les deux interpretations, un comite d’adjudication tranchait. Resultats 105 patients (39 femmes et 66 hommes), âges en moyenne de 60 ± 10 ans ont ete inclus de mai 2007 a fevrier 2008. Trente patients (29 %) etaient a risque moyen de CCR et 75 patients (71 %) a risque eleve. 90 % des VCC etaient excretees dans les 10 heures apres l’ingestion. La preparation du colon etait jugee bonne ou excellente dans 55 % des cas en VCC et 87 % des cas en coloscopie. La prevalence de polype significatif (≥ 6 mm) etait de 25 % pour la VCC et la coloscopie. Les performances de la VCC pour le diagnostic de polype ≥ 6 mm etaient : Sensibilite 55 %, Specificite 87 %, Valeur Predictive Positive 52 % et Valeur Predictive Negative 88 %. Chez les patients ayant une preparation bonne ou excellente, les performances etaient respectivement de 69 %, 84 %, 56 %, et 90 %. Apres adjudication des cas discordants, les performances etaient respectivement de 73 %, 93 %, 73 % et 93 %. Cinq evenements secondaires ont ete observes, en rapport avec la preparation (2 fois) ou la coloscopie (3 fois). La VCC a ete mieux appreciee que la coloscopie (echelle visuelle analogique : 8,80 vs 8,27 p Conclusion Ces resultats montrent l’excellente acceptabilite et tolerance de la VCC pour le depistage et le suivi de lesions neoplasiques colo-rectales. Cependant la preparation etait jugee insuffisante chez pres d’un malade sur deux et a conduit a modifier le protocole de preparation. De plus, les resultats apres adjudication confirment l’importance d’un apprentissage prealable a l’utilisation de cette technique. Malgre ces reserves, la combinaison d’une sensibilite de 73 % et d’une VPN de 93 % pour la detection de polypes de plus de 6 mm semble particulierement prometteuse.