When a source of upper GI bleeding is suspected, the test of choice for identifying and treating the bleeding lesion is endoscopy (
11). The timing of endoscopy remains a subject of debate, particularly for the patient who responds rapidly to volume replacement and has no further evidence of bleeding (
11). While it may seem intuitively obvious that endoscopy would improve care, randomized trials have indicated that diagnostic endoscopy does not reduce mortality, rebleeding rates, need for surgery or hospital stay (
12-
14). The overwhelming majority of patients (80%) with bleeding ulcers stop bleeding spontaneously, limiting the impact of early diagnosis (
14).
It is now clear that patients with continued or recurrent bleeding will benefit from urgent endoscopy (11). In this situation, endoscopy can improve morbidity and mortality. A consensus statement was issued by the National Institute of Health, in 1989, endorsing endoscopic therapy in patients with bleeding ulcers who are actively bleeding or at a high risk of rebleeding (15). A similar approach is advocated by the American Society for Gastrointestinal Endoscopy (16).
Summary of available data indicates that endoscopy has no beneficial effect on mortality and morbidity rates in Forrest’s III type ulcers. However, therapeutic endoscopy will improve mortality and morbidity rates in actively bleeding Forrest I ulcers. However, outcome of endoscopy in Forrest’s II ulcer, which is not actively bleeding but has a high risk of bleeding, remains controversial (8,9,11).
In the past decade, proton pump inhibitors have been commonly used in patients with peptic ulcer bleeding and the clinical outcome of omeprazole in the treatment of peptic ulcer has been excellent (9). Grosso et al compared perendoscopic prophylactic injection therapy with medical treatment by omeprazole infusion in Forrest's II ulcers (9). They evaluated the rate of rebleeding, hospital stay, need for transfusion, need for surgery and mortality in these two groups. They found no significant difference between the groups with respect to mortality and morbidity rates. However, at endoscopic follow-up 48 hours later, there were significantly more lesions with a higher risk of rebleeding (Forrest’s IIa and IIb) in the group treated with perendoscopic hemostasis. Their data suggested that omeprazole infusion is a valid alternative to injection treatment of non-bleeding visible vessels (9). In this study, better results were achieved in the medically treated group even though the other group was treated with perendoscopic injection plus medical the-rapy. One reason for this may be the fact that, unlike Grosso et al, we used both Omeprazole and Somatostatin in the medical treatment.
Some authors claim that perendoscopic prophylactic hemostasis should certainly be performed in patients with a visible vessel or sentinel clot (17-19). However, these data come from trials designed with combination therapies such as injection plus hemoclipping, bipolar thermal therapy plus laser hemostasis or two or more different combinations of these methods. The CURE Hemostasis Research Group has also reported better outcomes with combination therapy than monotherapy (11). In our study, perendoscopic prophylactic hemostasis was obtained by injection of sclerosing material as no other method of perendoscopic hemostasis was available in our endoscopy laboratory. Hence we were unable to evaluate the outcome of combination therapies in these patients. Injection of sclerosing material without combining any other method of hemostasis can stop active bleeding effectively but may have some harmful effects on the mucus layer and mucosal and submucosal structures (18,19). In non-bleeding ulcers with a high risk of bleeding (Forrest’s II), this mucosal destructive effect of injection therapy may be greater than its hemostatic effect.
Somatostatin has well defined effects on gastric secretion and gastroduodenal blood flow and this has led to the use of somatostatin in the treatment of upper gastrointestinal hemorrhage (20,21). Somatostatin was used in both groups in this study, which might have improved clinical outcomes. However, it is not possible to compare the two groups in terms of somatostatin effects because of the design of the trial.
Our results showed that the clinical outcome of medical therapy alone with somatostatin and omeprazole was better than perendoscopic injection therapy combined with same medical treatment. The rate of early rebleeding, number of blood units transfused and number of endoscopic procedures required was lower in the group treated without injection therapy. Moreover, ulcer improvement occurred much more rapidly and duration of hospital stay was shorter in this group.
There may be several reasons for these results. Firstly, it was not possible to use a combination of two or more perendoscopic prophylactic hemostasis methods and the injection of sclerosing material as a monotherapy was used. Secondly, medical treatment was given to both groups. The aim of the study was to evaluate the use of perendoscopic injection therapy in Forrest II ulcers rather than compare medical and perendoscopic treatments because it would be unethical to omit omeprazole and somatostatin for the purpose of this trial. Finally, the small number of patients may limit the conclusions to be drawn from this study.
In conclusion, the indication for perendoscopic prophylactic injection monotherapy in non-bleeding ulcers with a high risk of rebleeding should be reviewed in large population based, prospective, randomized trials.
){kind=link}
){kind=link}