Fibrin glue in the endoscopic treatment of fistulae and anastomotic leakages of the gastrointestinal tract.

BACKGROUND: Fistulae or leakages of anastomotic junctions of the gastrointestinal tract used to be an indication for surgery. However, patients often are severely ill and endoscopic therapeutic options have been suggested to avoid surgical intervention. PURPOSE: This is a retrospective analysis of fibrin glue application in the treatment of gastrointestinal fistulae or anastomotic leakages. AIM: The aim of this study was to investigate the value of fibrin glue in the treatment of gastrointestinal fistulae and leakages. METHODS: From September 1996 to November 2002, 52 patients with gastrointestinal fistulae or insufficiencies have been treated endoscopically including the use of fibrin glue (Tissucol Duo S®, Baxter, Unterschleissheim, Germany). Clinical data comprising concomitant therapies and results were analysed by chart review. RESULTS: Twenty-six lesions were located in the oesophagus or gastroesophageal junction, 4 in the stomach, 7 in the small intestine, 13 colorectal and 2 in the pancreas. The duration of treatment ranged from 12 to 1,765 days. Two to 81 ml fibrin glue (median 8.5) was used in 1-40 sessions (median 4). All patients received antibiotics; additional endoscopic options were frequently applied. Endoscopic therapy cured 55.7% patients (n = 29); 36.5% (n = 19) were cured with fibrin glue as sole endoscopic option. In 23.1% (n = 12), surgical intervention became necessary. Patients without major infectious complications tended to have a higher cure rate without surgery (87.5% vs. 50%). Eleven patients died (21.1%). CONCLUSION: Endoscopic therapy is a valuable option in the treatment of fistulae and anastomotic insufficiencies of the gastrointestinal tract. It usually is applied repeatedly. Fibrin glue is a mainstay of this procedure. Major infectious complications seem to define a subgroup of patients with poorer outcome.

Therapy insight: Prophylaxis of stress-induced gastrointestinal bleeding in critically ill patients.

Stress-induced gastrointestinal bleeding is associated with increased morbidity and mortality in critically ill patients. Within the past few decades, the incidence of stress-induced gastrointestinal bleeding has decreased. Prophylaxis of stress-induced gastrointestinal bleeding, which is aimed at preventing morbidity and mortality, has to be achieved with as few adverse effects as possible. Data indicate that not all critically ill patients need prophylaxis for stress-induced gastrointestinal bleeding. The main risk factors associated with clinically important hemorrhage are mechanical ventilation for >48 h, and coagulopathy (thrombocyte count <50/nl, partial thromboplastin time (PTT) >2 times the upper limit of the normal range, international normalized ratio (INR) >1.5). Ranitidine is more effective than sucralfate for the prevention of clinically important bleeding. Immediate-release omeprazole is as effective as cimetidine, and is more efficient at increasing the intragastric pH. As yet, however, there is no firm evidence that any of the drugs used for prophylaxis of stress-induced gastrointestinal bleeding in critically ill patients decrease mortality or the length of hospital stay. When to stop prophylaxis is decided on clinical grounds rather than on the basis of data from clinical studies.

Association of antibodies to exocrine pancreas with subtypes of Crohn's disease.

OBJECTIVES: The Vienna classification of Crohn's disease (CD) subdivides patients according to their age at diagnosis, disease location, and disease behaviour. This study tested whether antibodies against exocrine pancreatic tissue (PAbs) are associated with Vienna classification subtypes of CD. METHODS: PAbs were examined by indirect immunofluorescence in sera of patients with CD (n = 208), or ulcerative colitis (n = 47), and in normal controls (n = 50). Presence of PAbs was compared to the Vienna classification, disease activity, and drug therapy. RESULTS: PAbs were present in 60 CD patients (28.8%), but in no patient with ulcerative colitis or in controls (P < 0.001). There was no significant association between PAbs and disease activity, or drug therapy. Complete classification into Vienna subtypes was possible in 168 patients. There was significant heterogeneity in PAb prevalence in these CD subtypes (P < 0.01). Variation was found in the behaviour category (B1, non-stricturing, non-penetrating CD: 14.0% PAb-positive; B2, stricturing CD: 41.7%; B3, penetrating CD: 31.5%; P = 0.02), but not age or location categories. PAbs were associated with long disease duration (P = 0.04). CONCLUSION: This study confirms that PAbs are highly specific for CD. The small differences in their prevalence in CD subtypes do not suggest that PAb detection is useful in discrimination of CD phenotypes.

Comparison of inpatient and outpatient upper gastrointestinal haemorrhage.

BACKGROUND AND AIMS: Inpatients developing upper gastrointestinal (GI) haemorrhage are at increased risk of death. This study was performed to elucidate differences in inpatients and outpatients. PATIENTS/METHODS: Three hundred and sixty-two patients who needed esophagogastroduodenoscopy for upper GI bleeding were identified from endoscopy charts. Patients' characteristics, bleeding parameters, clinical presentation, pre-existing medication, and laboratory data were compared between patients who were admitted because of upper GI bleeding and patients who developed bleeding while in hospital for other reasons. RESULTS/FINDINGS: Hospital mortality was 39.0% in inpatients vs. 11.1% in outpatients (p<0.01). Death due to bleeding was observed in 9.5% of inpatients vs. 2.5% of outpatients (p<0.01). Whereas peptic ulcer was the most common source of bleeding in both, variceal bleeding was the most common cause of death because of haemorrhage in both. Recurrent bleeding was associated with mortality in outpatients (p<0.001), but not in inpatients (p=0.11). Rates of bleeding recurrence and need for surgery was similar in both groups. Inpatients suffered more often from renal disease, pulmonary disease, diabetes mellitus, coagulopathy, or immunosuppression, and were treated more frequently with acetylsalicylic acid, glucocorticoids and heparin. The frequency of pre-existing disease was higher in inpatients. INTERPRETATION/CONCLUSION: Higher mortality after GI bleeding in inpatients than in outpatients is due to a generally higher prevalence of co-morbidity rather than a single or a few risk factors.

Optimising the diagnostic value of anti-Saccharomyces cerevisiae-antibodies (ASCA) in Crohn's disease.

BACKGROUND/AIMS: Anti-Saccharomyces cerevisiae-antibodies (ASCA) are used to discriminate Crohn's disease (CD) from ulcerative colitis (UC). ASCA tests are not standardised and different methods for ASCA detection exist. This study was undertaken to compare ASCA tests and to clarify their diagnostic value. PATIENTS/METHODS: One hundred and two sera from CD patients, 53 from UC patients, and 50 sera from normal controls were examined for ASCA IgA and ASCA IgG using four different ELISA (Aesku.lab, Inova, Euroimmun, and Medipan) and indirect immunofluorescence (Euroimmun). Sensitivity, specificity, and positive and negative predictive values (PPV and NPV) for CD in this population were determined. Agreement between tests was expressed by kappa statistics. RESULTS/FINDINGS: The presence of either ASCA IgA or ASCA IgG had a PPV between 77 and 88%. Only the combined presence of ASCA IgA and ASCA IgG had a specificity which was constantly above 90%. Sensitivity, specificity, PPV, and NPV were then: Inova: 53, 97, 95, and 68%; Euroimmun ELISA: 46, 97, 94, and 66%; Aesku.lab: 50, 97, 94, and 66%; Medipan: 30, 98, 94, and 59%; indirect immunofluorescence: 51, 97, 94, and 66%. Agreement between ELISA test results of the Aesku.lab, Euroimmun, Medipan IgA and Inova systems was good (kappa: 0.63-0.79); between the Medipan IgG ELISA or indirect immunofluorescence and the others, it was lower (kappa: 0.33-0.6). If both ASCA IgA and ASCA IgG were detected by indirect immunofluorescence as well as ELISA, specificity for CD increased to >99% at a sensitivity of 23-38%. INTERPRETATION/CONCLUSIONS: The combined detection of ASCA IgA and ASCA IgG by indirect immunofluorescence as well as ELISA may optimise the discrimination of CD from UC.

[Detection of lymph node metastases of malignant melanoma by palpation and ultrasound]. / Palpatorische und sonographische Detektion von Lymphknotenmetastasen bei lokal fortgeschrittenem malignen Melanom.

BACKGROUND AND PURPOSE: Early detection of metastases of malignant melanoma has therapeutic implications. The aim of this study was to evaluate palpation and ultrasound examination in the diagnostics of lymph node metastases in locally advanced melanoma. PATIENTS AND METHODS: 83 patients suffering from melanoma (Clark level IV or V) were examined for lymph node metastases by palpation and sonography. Findings were compared to histopathologic results after lymph node extirpation if available or the findings at the next follow-up visit. RESULTS: Lymph node metastases were confirmed histopathologically in 14 patients at the first study visit, in three others at the control visit. Sensitivity, specificity, positive and negative predictive values of palpation for the detection of metastases or suspicious nodes with increasing volume at follow-up in this population were 65%, 81%, 48%, and 89%, and of ultrasound 100%, 66%, 45%, and 100%, respectively. CONCLUSION: Sonography of lymph nodes should be included as a standard procedure in the detection of metastases of locally advanced malignant melanoma.

Association of perinuclear antineutrophil cytoplasmic antibodies and anti-Saccharomyces cerevisiae antibodies with Vienna classification subtypes of Crohn's disease.

BACKGROUND: The Vienna classification of Crohn's disease (CD) subdivides patients according to their age at diagnosis (A), disease location (L), and disease behavior (B). AIM The aim of this study was to test whether perinuclear antineutrophil cytoplasmic antibodies (pANCAs) or anti-Saccharomyces cerevisiae antibodies (ASCAs) correlate to subtypes of CD according to this classification. METHODS: pANCA, ASCA-immunoglobulin (Ig) A, and ASCA-IgG were detected by indirect immunofluorescence in 120 sera of patients with CD and compared with their Vienna classification. RESULTS: Patients with diagnosis of CD at an age of 40 years or older (A2) were more frequently pANCA-positive than those whose disease started at a younger age (A1) (19% vs. 5%, p < 0.05). pANCA-positive patients almost exclusively belonged to the nonpenetrating disease groups (B1 and B2); only one patient had penetrating disease (B3) (B3 vs. non-B3: p = 0.02). There was a cluster of pANCA-positive patients within the A1 B2 L3 subgroup (B2: stricturing disease; L3: ileocolonic involvement). Only 7.5% of ASCA-IgA-positive patients presented with CD that was limited to the colon (L2). ASCA-IgA and ASCA-IgG were more frequently found in patients with upper gastrointestinal disease (L4) (ASCA-IgA: 66.7% vs. 31.5% in non-L4, p < 0.05; ASCA-IgG: 44.4% vs. 27.9%, p < 0.05). CONCLUSION: pANCA and ASCA may be useful in predicting subtypes of CD. They therefore may be helpful in developing subtype-specific disease management strategies.