Search

Capsule endoscopy (CE) is a new modality for obscure digestive bleeding (OBD), but little is known about its influence on management and outcome.Fifty-six patients (male/female 26/30; mean age 63 yr) with ODB, and negative upper and lower gastrointestinal (GI) endoscopy were included in this multicenter study. The diagnostic yield of CE was compared to three other tests (OT: push enteroscopy, enteroclysis, angiography), and patients were followed up for at least 6 months. Parameters were analyzed that led to major management changes such as surgical or endoscopic intervention or specific medical therapy, as well as their correlation to further bleeding.CE had a diagnostic yield higher than OT (68% vs 38%). Major management changes and an improvement in bleeding activity were observed in 21 and 44 patients, respectively. The number and type of positive findings on CE were associated with significant management changes (p0.05). The number of positive findings detected by CE as well as by OT correlated with further bleeding episodes (p0.05). However, clinical parameters (lowest hemoglobin (Hb) value, number of blood transfusions) were also significantly associated with outcome. Diagnoses of specific diseases (tumor, Crohn, NSAID ulcer) by CE led to a favorable outcome in 64% of cases, whereas negative CE cases were associated with no further bleeding in 80%.CE helps with management decisions and can replace other more complex and risky standard tests. Nevertheless, clinical parameters are equally important for predicting further bleeding and should also be used to decide on further management.

Endoscopic appendectomy in a case of appendicular intussusception due to endometriosis, mimicking a cecal polyp

Despite improvements of diagnostic modalities differentiation between benign and malignant hilar strictures remains a challenge. Hilar neoplasia requires preoperative tissue diagnosis to avoid risk of inappropriate extensive surgery. This is commonly attempted using various techniques at ERCP, which have variable sensitivity and accuracy. We used endosonography-guided fine-needle aspiration (EUS-FNA) for the preoperative diagnosis of hilar cholangiocarcinoma (HC).Prospective evaluation of 44 patients (31 male, mean age: 59 yr) with strictures at the liver hilum were diagnosed by CT and/or ERCP. All were suspicious of HC but had inconclusive tissue diagnosis. They underwent EUS-FNA with linear echo endoscope and 22 gauge needles.Adequate material was obtained in 43 of 44 patients. Cytology revealed HC in 26 and other malignancies in 5 patients; 12 had benign results: sclerosing cholangitis (n = 4), primary sclerosing cholangitis (n = 4), inflammation (n = 3), sarcoid-like lesion (n = 1). There were no significant differences in age, lesion size, or echo features among patients with adenocarcinomas, other malignancies, or benign lesions. Thirty-two patients underwent surgery, 2 had autopsy, 10 were followed up clinically. Four of the benign results were false negatives. No complications occurred. Accuracy, sensitivity, and specificity were 91%, 89%, and 100%, respectively. EUS and EUS-FNA changed preplanned surgical approach in 27 of 44 patients.These results suggest that EUS-FNA is of value as a new, less-invasive approach for tissue diagnosis of hilar strictures of unknown cause. It was technically feasible without significant risks, when other diagnostic tests were inconclusive and was able to change preplanned management in about half of the patients.

OBJECTIVES: The clinical value of endoscopic ultrasound-guided fine needle aspiration (EUS-FNA) of pancreatic lesions is uncertain in patients with normal parenchyma and chronic pancreatitis. The aim of this study was to analyze the diagnostic yield and influence of EUS-FNA on the clinical management of patients with pancreatic lesions, in the presence (CP) or absence (NP) of chronic pancreatitis. METHODS: A total of 207 consecutive patients with NP (n = 133) and CP (n = 74) were examined using linear array echo endoscopes for the procedure and 22-gauge needles. RESULTS: Adequate specimens were obtained from 200 lesions. A correct final diagnosis was established at histology (n = 108), bacteriology (n = 9), and clinical follow-up (n = 83). Cytology gave 17 false-negative EUS-FNA results (overall sensitivity: 85%). In patients with NP, 60 solid adenocarcinomas were detected, 32 other malignancies, and 38 benign lesions, with 11 false-negative results (sensitivity: 89%). In patients with CP, only seven of 13 malignancies (all solid adenocarcinomas) were identified using FNA (sensitivity: 54%). Overall, malignancy was identified in 116 patients, 32 of whom (27%) had lesions other than primary solid adenocarcinomas. Management was altered in 25 of these patients, which changed the surgical approach in 21%. EUS-FNA influenced the therapeutic approach in 44% of the total patient group. CONCLUSIONS: EUS-FNA was especially useful in patients with a focal pancreatic lesion with normal parenchyma. Its sensitivity in patients with CP was unacceptably low, and resection of the tumor using standard surgical techniques was still usually required to confirm the correct diagnosis. Diagnostic EUS-FNA influenced clinical management in nearly half of patients.

OBJECTIVE: Mediastinal lymphadenopathy (ML) is a cause for concern, especially in patients with previous malignancy. The investigation of choice is thoracic CT with a variable sensitivity and specificity requiring tissue diagnosis. We used endoscopic ultrasound-guided fine-needle aspiration (EUS-FNA) for cytodiagnosis of ML in patients with and without previous malignancy. The cause, distribution of lesions, and incidence of second cancers were investigated. METHODS: Linear echoendoscopes and 22-gauge needles for cytology were used for EUS-FNA. A cytological diagnosis of malignancy was accepted, and histology or consistent follow-up of at least 9 months confirmed benign results. RESULTS: One hundred fifty-three patients underwent EUS-FNA between November 1997 and November 1999 (mean age, 60 yr; range, 13-82 yr; 105 men). Cytology was adequate in 150 patients. Final diagnosis was malignancy in 84 and benign in 66 patients (sensitivity, specificity, and diagnostic accuracy: 92%, 100%, 95%, respectively). In 101 patients without previous cancer cytology identified 48 malignant (lung, 41; extrathoracic, 7) and 51 benign lesions (inflammation, 35; various, 9; sarcoidosis, 7) (sensitivity, specificity, accuracy: 88%, 100%, 94%). Fifty-two patients had prior malignancy, mostly in extrathoracic sites. Cytology revealed recurrences in 21 patients, second cancer in 9 and benign lesions in 21 patients (inflammatory, 11; sarcoidosis, 8; tuberculosis, 1; abscess, 1) (sensitivity, specificity, accuracy: 97%, 100%, 98%). CONCLUSIONS: In patients without previous cancer malignant ML originates from the lung >80%. In those with previous malignancy recurrence of extrathoracic sites is the major cause. Benign lesions and treatable second cancers occur in a significant frequency, emphasizing the need for tissue diagnosis. EUS-FNA is a safe and minimally invasive alternative for cytodiagnosis in the mediastinum.

Organ preserving pancreatic resections are considered whenever malignant disease is ruled out. In tumors of low malignant potential such as cystadenomas and neuroendocrine tumors, the diagnosis is rarely established preoperatively. We studied the feasibility of cytodiagnosis using endosonography-guided fine-needle aspiration in determining the operative approach.A total of 78 patients (16 female, 62 male; mean age 61.4 yr, range 31-82 yr) with focal pancreatic lesions underwent EUS-FNA. Final diagnosis was confirmed by histology, cytology, or clinical follow up (9 months). Patients with tumors of low malignant potential were managed by customized pancreatic resections.Final diagnosis was malignant tumors in 36 patients, tumors of low malignant potential in nine (six, neuroendocrine, two, borderline mucinous cystadenomas, one, borderline adenocarcinoma), and benign in 31 (two inadequate smears). No complications occurred. With six false-negative and no false-positive results, the accuracy, sensitivity, specificity, and positive and negative predictive values were 92%, 84%, 100%, 100%, and 86%, respectively. Five patients with low malignant tumors underwent duodenum-preserving pancreatic head resection, three mid segment resection, and one pylorus-preserving pancreatoduodenectomy.EUS-FNA is useful in the preoperative cytodiagnosis of pancreatic tumors of low malignant potential. It extends the indication for organ-preserving pancreatic resections and avoids the unnecessary sacrifice of adjacent organs.