Ağustos 2010

27-31

Emrah ALPER¹, Behlül BAYDAR², Cem ÇEKİǹ, Fatih ASLAN¹, Serdar AKÇA², Belkıs ÜNSAL¹

¹Atatürk Training and Research Hospital, İzmir
²Memorial Antalya Hastanesi, Antalya

Amaç: Non variseal üst gastrointestinal sistem kanaması dünyada acil hastaneye yatış nedenleri arasında yaygındır ve hala yüksek mortalite oranlarına neden olur. Bu çalışmada, üst gastrointestinal sistem kanamasına neden olan lezyonlar Forrest sınıflamasına gore değerlendirildi ve bu lezyonlarla tekrar kanama arası ilişki açıklandı. Yöntem ve Gereç: Bu çalışma, tersiyer eğitim ve araştırma hastanesinde retrospektif olarak yapılmıştır. Kasım 2005 ile Mayıs 2009 arasında, son 24 saatte üst gastrointestinal sistem kanaması gelişen ve hastaneye başvuran hastalar çalışmaya alındı. Binaltıyüzkırkyedi hastanın verileri analiz edildi ve 1342 hastanın verileri değerlendirildi. Bulgular: Lezyonların dağılımı: 96?sı (%7.15) özofagusta, 552?si (%41.13) midede, 523?ü (%38.97) duodenumda ve 171?i (%12.74) birden çok bölgede yerleşmiştir. Ayrıca 34 (%2.53) lezyon Forrest Ia, 192?si (%14.31) Forrest Ib, 85?i (%6.33) Forrest IIa, 121?i (%9.02) Forrest IIb, 210?u (%15.65) Forrest IIc ve 700?ü (%52.16) Forrest III olarak sınışanmıştır. Sonuç: Sonuç olarak, non-variseal üst gastrointestinal sistem kanamalarının Forrest sınıflamasına göre değerlendirilmesinde, Forrest Ia, Ib ve IIa gruplarında tekrar kanama oranı, diğerleriyle kıyaslandığında daha yüksekti. Tekrar kanama riskini belirlemek için endoskopik değerlendirmenin ilk 12 saat içinde Forrest sınıflamasına göre yapılmasını önerebiliriz.

Tekrar kanama, üst gastrointestinal kanama, Forrest sınıflaması

Non-variceal upper gastrointestinal (GI) system bleeding is among the common reasons for urgent hospitalization worldwide and still causes high rates of mortality (1-7). The incidence of this category of bleeding ranges between 50/100,000 and 150/100,000 (8). While the bleeding ceases spontaneously in 80% of the patients, endoscopic treatment is required for the remaining 20%. Even though endoscopic treatment methods are effective in 90% of patients (9), bleeding after endoscopic treatment accounts for mortality in 20% of patients. A hemoglobin level <6 g/dl at the time of first admission, the location of ulcer at the bulbus (10, 11), a large-sized ulcer (11-14), hemodynamic instability, coexisting illnesses (1315), and active bleeding or signs of bleeding on endoscopy (11, 13, 16-19) determine the risk of rebleeding. The continuation of medical treatment with proton pump inhibitors(PPIs) after endoscopic treatment lowers the risk of rebleeding considerably (18, 20-24).
In the present study, the lesions causing upper GI bleeding were evaluated according to Forrest classification (25), and the relationship between these lesions and rebleeding was elucidated.

In the present study, a significant increase in the risk of rebleeding in lesions with active bleeding or visible vessels according to the Forrest classification was shown in the patients with upper GI bleeding, In the literature, the rates of rebleeding and mortality have been reported to be 12-39% and 3.5-35.7% following treatment, respectively, whereas the number of cases requiring surgery has been reported to be 2.8-14.2% (1, 2, 7, 12, 14, 21, 2628). In the present study, the rate of rebleeding and mortality was 6.86%. This is lower than the rates reported in other studies. (2-7, 14, 15) It has been suggested that this is due to the inability to identify the bleeding site in cases of excessive bleeding, exclusion of patients from the study who could not be categorized according to the Forrest classification, performing endoscopic intervention in patients in Forrest class IIa, the presence of variability in distribution according to the Forrest classification, and intensive PPI infusion therapy.

The endoscopic classification of upper GI bleeding was first reported by Forrest et al. in 1974 (25). The risk determination based on the Forrest classification is currently one of the most commonly used methods of evaluation. Methods evaluating clinical findings and laboratory parameters of patients have also been used. In the prospective study by Kim et al. (7) comparing five different scoring systems in patients with nonvariceal bleeding, it was demonstrated that the Forrest classification was more specific and had a higher positive predictive value when compared to the other scoring systems with respect to rebleeding and mortality risk determination (7). In a study involving 239 patients, the rate of bleeding was 14.6%, whereas the mortality rate was 8.4%, and it was reported that endoscopic evaluation was necessary for risk determination in patients with upper GI bleeding (7). In the same study, the sensitivity of Forrest classification was found to be 71.4% for rebleeding and 85% for mortality. In a study conducted by Manguso et al. (1), analysis of endoscopic findings of 142 patients with Forrest class Ia and Ib showed that rebleeding occurred in 12% of patients, mortality throughout the hospitalization was 5.6%, and the rate of patients who underwent surgery due to inability to stop bleeding was 2.8%. In upper GI bleeding associated with peptic ulcers, despite the importance of clinical factors, certain studies have reported that the Forrest classification was the principal determining factor (17). In the present study, it was determined that as the severity of factors of active bleeding increased, the risk of rebleeding also increased according to the Forrest classification. It was observed that rebleeding developed in 41.1%, 22.9% and 12.9% of the patients classified as Forrest Ia, Ib and Ic, respectively. On the other hand, the rate of rebleeding in Forrest class III was 1.7%.

In a study conducted by Chung et al. (12), the Forrest classification and location and size of the lesion were evaluated together. Based on univariate analyses, rebleeding was found to be associated with Forrest class Ia, ulcer sizes >2 cm and lesions located in the stomach (12). In a study involving 738 patients who had been evaluated for rebleeding and mortality using multivariate analyses, liver cirrhosis, recent surgery, low systolic blood pressure, hematemesis, Forrest classification, and localization and size of the ulcer were investigated. According to the data, four classes were determined, and the Forrest classification was reported to be the best indicator for early rebleeding (within the first 48 hours). The sensitivity in the first 48 hours was 90%, whereas it was 65% after 48 hours (3). In the present study, the risk of rebleeding was higher in ulcers located in the esophagus and duodenum, and in ulcers classified as Forrest Ia, Ib and IIa, and it was determined that rebleeding and mortality occurred in 92.3% of the patients during the first 48 hours.

In conclusion, evaluation of non-variceal upper GI bleeding according to the Forrest classification demonstrated that the rate of rebleeding in the Forrest Ia, Ib and IIa groups was higher when compared to the others. It is suggested that endoscopic evaluation within the first 12 hours should be performed according to the Forrest classification in order to determine the risk of rebleeding.

The present study was a retrospective study conducted at the Izmir Ataturk Training and Research Hospital, a tertiary training and research hospital, involving patients with non-variceal upper GI bleeding who had undergone upper GI endoscopy and were being treated with medical and endoscopic techniques. Patients who were admitted to the outpatient clinic and emergency units of the hospital between November2005 and May 2009 with upper GI bleeding developed within the last 24 hours were included in the study. Upper GI endoscopy was performed in all patients within 12 hours of admission and the non-variceal causes of bleeding were identified. All patients (except for mortalities) were hospitalized and followed for at least 5 days and treated with 80 mg pantoprazole intravenous (i.v.) push, followed by an i.v. infusion at 8 mg/hour for at least 72 hours. Data were obtained through evaluation of records of the endoscopy laboratory and investigation of patient documents from the hospital database. Biopsies were obtained for histologic sampling from patients with a single gastric ulcer and from patients in whom the endoscopic evaluation demonstrated lesions in the esophagus and stomach suggestive of malignancy with no contraindication for biopsy. Patients with variceal bleeding, cirrhotic patients with non-variceal bleeding, patients whose lesions could not be evaluated due to excessive bleeding, patients with bleeding following a recent GI surgical procedure, and patients with histologically detected malignancies were excluded from the study. Demographic data, patient history and treatments, and findings of biochemical and hematology laboratories were obtained from patient records; localization of lesions, endoscopic levels of lesions according to the Forrest classification, and endoscopic treatment and their types were analyzed from endoscopy laboratory files.
Bleeding was considered as rebleeding or continuous bleeding when it could not be stopped by endoscopic or medical treatment, and in cases of recurrence of bleeding, death due to bleeding or bleeding requiring surgical intervention, hematemesis following endoscopic or medical treatment, fresh bleeding from a nasogastric tube, a decrease in the hemoglobin level >2 g/dl within 24 hours, hypotension (<90/60 mmHg), tachycardia (>130/min), and the development of signs of shock. Signs of active bleeding were observed in this group of patients following a second upper GI endoscopy or operations for surgical treatment. Attempts were made to treat patients with rebleeding endoscopically or surgically. Statistical analyses were performed using SPSS 13.0 for Windows (SPSS Inc., Chicago, IL, USA). Patient characteristics and variables were analyzed by chi-square test and Student?s t test. Odds ratios were also calculated.

The data of 1647 patients fulfilling the study criteria, who were admitted with complaints of upper GI bleeding within 24 hours and who underwent upper GI endoscopy, were analyzed. Three hundred five patients were excluded from the study according to the exclusion criteria; thus, the data for 1342 patients were evaluated. The mean age of the patients was 61 years (range, 17-91 years); 755 females (56.26%) and 587 males (43.74%) were evaluated. Seven hundred twenty (53.2%) of the patients had a history of non-steroidal anti-inflammatory drug (aspirin; Bayer AG, Leverkusen, Germany) use within the past 7 days, while 52 (3.9%) of the patients had a history of anticoagulant drug use. Ninety-six of the patients had a history of upper GI bleeding. The mean hemoglobin level during the first admission was 10.1 g/dl (Table 1).

Endoscopic evaluation was performed on 242 (18.03%) and 1100 (81.07%) patients before and after PPI treatment, respectively. In the 265 patients with lesions classified as Forrest Ia, Ib and IIa, sclerotherapy with 1/10,000 adrenalin was performed, while endoscopic therapy with sclerotherapy and clips was performed in 5 patients (Table 1). All patients, except the ones who died or underwent surgery, received medical treatment with a pantoprazole infusion (8 mg/hour) for at least 3 days.

No significant difference was found between patients with rebleeding and patients without bleeding with respect to age, gender and the etiology of bleeding (p<0.05). The mean hemoglobin level of patients with bleeding was 9.2 g/dl on admission, whereas it was 10.6 g/dl in those without bleeding. The difference between the two groups was statistically significant (p>0.05).

Distribution of the lesions was as follows: 96 (7.15%) esophagus, 552 (41.13%) stomach, 523 (38.97%) duodenum, and 171 (12.74%) in multiple areas. It was also demonstrated that 34 (2.53%) of the lesions were classified as Forrest Ia, 192 (14.31%) as Forrest Ib, 85 (6.33%) as Forrest IIa, 121 (9.02%) as Forrest IIb, 210 (15.65%) as Forrest IIc, and 700 (52.16%) as Forrest III (Table 2).

No rebleeding was observed in 1250 (93.14%) of the patients who underwent medical and endoscopic therapy during the 5-day follow-up and treatment period. Rebleeding and mortality associated with rebleeding was observed in 92 (6.86%) of the patients. Mortality occurred in 6 (0.45%) patients; 3 of the patients died before surgery -1 during surgery and 2 during the postoperative period. Rebleeding and mortality was observed in 85 (92.3%) of the patients within the first 48 hours. Of the lesions causing rebleeding and mortality, 14 (15.21%) were classified as Forrest Ia, 44 (47.83%) as Forrest Ib, 11 (11.96%) as Forrest IIa, 7 (7.61%) as Forrest IIb, 4 (4.35%) as Forrest IIc, and 12 (13.04%) as Forrest III (Table 3).

Evaluation according to endoscopic severity demonstrated that the rates of rebleeding were 41.18%, 22.92%, 12.94%, 5.79%, 1.91% and 1.71% in Forrest class Ia, Ib IIa, IIb, IIc, and III lesions, respectively. Odds ratios were 11.03, 6.82, 2.15, 0.82, 0.23 and 0.12 in Forrest class Ia, Ib, IIa, IIb, IIc, and III lesions, respectively (Table 2).

In the present study, a significant increase in the risk of rebleeding in lesions with active bleeding or visible vessels according to the Forrest classification was shown in the patients with upper GI bleeding, In the literature, the rates of rebleeding and mortality have been reported to be 12-39% and 3.5-35.7% following treatment, respectively, whereas the number of cases requiring surgery has been reported to be 2.8-14.2% (1, 2, 7, 12, 14, 21, 2628). In the present study, the rate of rebleeding and mortality was 6.86%. This is lower than the rates reported in other studies. (2-7, 14, 15) It has been suggested that this is due to the inability to identify the bleeding site in cases of excessive bleeding, exclusion of patients from the study who could not be categorized according to the Forrest classification, performing endoscopic intervention in patients in Forrest class IIa, the presence of variability in distribution according to the Forrest classification, and intensive PPI infusion therapy.

The endoscopic classification of upper GI bleeding was first reported by Forrest et al. in 1974 (25). The risk determination based on the Forrest classification is currently one of the most commonly used methods of evaluation. Methods evaluating clinical findings and laboratory parameters of patients have also been used. In the prospective study by Kim et al. (7) comparing five different scoring systems in patients with nonvariceal bleeding, it was demonstrated that the Forrest classification was more specific and had a higher positive predictive value when compared to the other scoring systems with respect to rebleeding and mortality risk determination (7). In a study involving 239 patients, the rate of bleeding was 14.6%, whereas the mortality rate was 8.4%, and it was reported that endoscopic evaluation was necessary for risk determination in patients with upper GI bleeding (7). In the same study, the sensitivity of Forrest classification was found to be 71.4% for rebleeding and 85% for mortality. In a study conducted by Manguso et al. (1), analysis of endoscopic findings of 142 patients with Forrest class Ia and Ib showed that rebleeding occurred in 12% of patients, mortality throughout the hospitalization was 5.6%, and the rate of patients who underwent surgery due to inability to stop bleeding was 2.8%. In upper GI bleeding associated with peptic ulcers, despite the importance of clinical factors, certain studies have reported that the Forrest classification was the principal determining factor (17). In the present study, it was determined that as the severity of factors of active bleeding increased, the risk of rebleeding also increased according to the Forrest classification. It was observed that rebleeding developed in 41.1%, 22.9% and 12.9% of the patients classified as Forrest Ia, Ib and Ic, respectively. On the other hand, the rate of rebleeding in Forrest class III was 1.7%.

In a study conducted by Chung et al. (12), the Forrest classification and location and size of the lesion were evaluated together. Based on univariate analyses, rebleeding was found to be associated with Forrest class Ia, ulcer sizes >2 cm and lesions located in the stomach (12). In a study involving 738 patients who had been evaluated for rebleeding and mortality using multivariate analyses, liver cirrhosis, recent surgery, low systolic blood pressure, hematemesis, Forrest classification, and localization and size of the ulcer were investigated. According to the data, four classes were determined, and the Forrest classification was reported to be the best indicator for early rebleeding (within the first 48 hours). The sensitivity in the first 48 hours was 90%, whereas it was 65% after 48 hours (3). In the present study, the risk of rebleeding was higher in ulcers located in the esophagus and duodenum, and in ulcers classified as Forrest Ia, Ib and IIa, and it was determined that rebleeding and mortality occurred in 92.3% of the patients during the first 48 hours.

In conclusion, evaluation of non-variceal upper GI bleeding according to the Forrest classification demonstrated that the rate of rebleeding in the Forrest Ia, Ib and IIa groups was higher when compared to the others. It is suggested that endoscopic evaluation within the first 12 hours should be performed according to the Forrest classification in order to determine the risk of rebleeding.

1. Manguso F, Riccio E, Bennato R, et al. In-hospital mortality in non-variceal upper gastrointestinal bleeding Forrest 1 patients. Scand J Gastroenterol 2008; 43: 1432-41.

2. Bourienne A, Pagenault M, Heresbach D, et al. Multicenter prospective study of prognostic factors of gastroduodenal ulcer hemorrhages. Reevaluation of clinical and endoscopic factors in the era of endoscopic hemostasis. Gastroenterol Clin Biol 2000; 24: 193-200. Abstract.

3. Guglielmi A, Ruzzenente A, Sandri M, et al. Risk assessment and prediction of rebleeding in bleeding gastroduodenal ulcer. Endoscopy 2002; 34: 778-86.

4. Pundzius J. Clinical and endoscopic signs for the prediction of recurrent bleeding from gastroduodenal ulcers. Eur J Surg 1994; 160: 689-92.

5. Garripoli A, Mondardini A, Turco D,et al. Hospitalization for peptic ulcer bleeding: evaluation of a risk scoring system in clinical practice. Dig Liver Dis 2000; 32: 577-82.

6. Bourienne A, Pagenault M, Heresbach D, et al. Multicenter prospective study of prognostic factors of gastroduodenal ulcer hemorrhages. Reevaluation of clinical and endoscopic factors in the era of endoscopic hemostasis. Gastroenterol Clin Biol 2000; 24: 193-200. Abstract.

7. Kim BJ, Park MK, Kim SJ, et al. Comparison of scoring systems for the prediction of outcomes in patients with nonvariceal upper gastrointestinal bleeding: a prospective study. Dig Dis Sci 2008 Dec 23. [Epub ahead of print].

8. Conrad SA. Acute upper gastrointestinal bleeding in critically ill patients: causes and treatment modalities. Crit Care Med 2002; 30: S365-8.

9. Laine L, Peterson W. Medical progress: bleeding peptic ulcer. N Engl J Med 1994; 331: 717-27.

10. Segal F, Prolla JC, Maguilnik I, Wolff FH. Clinical and endoscopic aspects in the evolution of patients with bleeding peptic ulcer--a cohort study. Arq Gastroenterol 2000; 37: 162-7.

11. Khadzhibaev AM, Malikov IuR, Kholmatov RM, et al. The role of endoscopy in diagnosis and treatment of gastroduodenal bleedings. Khirurgiia (Mosk). 2005; (4):24-7. (Abstract).

12. Chung IK, Kim EJ, Lee MS, et al. Endoscopic factors predisposing to rebleeding following endoscopic hemostasis in bleeding peptic ulcers. Endoscopy 2001; 33: 969-75.

13. Elmunzer BJ, Young SD, Inadomi JM, et al. Systematic review of the predictors of recurrent hemorrhage after endoscopic hemostatic therapy for bleeding peptic ulcers. Am J Gastroenterol 2008; 103: 2625-32.

14. de Manzoni G, Catalano F, Festini M, et al. Acute hemorrhage caused by duodenal ulcer. Results of endoscopic treatment of the first bleeding episode and of recurrences. Ann Ital Chir 2002; 73: 387-94. (Abstract).

15. van Leerdam ME. Epidemiology of acute upper gastrointestinal bleeding. Best Pract Res Clin Gastroenterol 2008; 22: 209-24.

16. Lau JY, Chung SC, Leung JW, et al. The evolution of stigmata of hemorrhage in bleeding peptic ulcers: a sequential endoscopic study. Endoscopy 1998; 30: 513-8.

17. Zaragoza AM, Tenías JM, Llorente MJ, Alborch A. Prognostic factors in gastrointestinal bleeding due to peptic ulcer: construction of a predictive model. J Clin Gastroenterol 2008; 42: 786-90.

18. Travis AC, Wasan SK, Saltzman JR. Model to predict rebleeding following endoscopic therapy for non-variceal upper gastrointestinal hemorrhage. J Gastroenterol Hepatol 2008; 23: 1505-10.

19. Guglielmi A, Ruzzenente A, Sandri M, et al. Risk assessment and prediction of rebleeding in bleeding gastroduodenal ulcer. Endoscopy 2002; 34: 778-86.

20. Keyvani L, Murthy S, Leeson S, Targownik LE. Pre-endoscopic proton pump inhibitor therapy reduces recurrent adverse gastrointestinal outcomes in patients with acute non-variceal upper gastrointestinal bleeding. Aliment Pharmacol Ther 2006; 24: 1247-55.

21. Lau JY, Sung JJ, Lee KK, et al. Effect of intravenous omeprazole on recurrent bleeding after endoscopic treatment of bleeding peptic ulcers. N Engl J Med 2000; 343: 310-6.

22. Lin HJ, Lo WC, Lee FY,et al. A prospective randomized comparative trial showing that omeprazole prevents rebleeding in patients with bleeding peptic ulcer after successful endoscopic therapy. Arch Intern Med 1998; 158: 54-8.

23. Tseng GY, Lin HJ, Lin HY, et al. The influence of intravenous omeprazole on intragastric pH and outcomes in patients with peptic ulcer bleeding after successful endoscopic therapy--a prospective randomized comparative trial. Hepatogastroenterology 1999; 46: 2183-8.

24. McCarthy DM. Management of bleeding peptic ulcer: current status of intravenous proton pump inhibitors. Best Pract Res Clin Gastroenterol. 2004; 18(Suppl): 7-12.

25. Forrest JAH, Finlayson NDC, Shearman DJ. Endoscopy in gastrointestinal bleeding. Lancet 1974; 2: 394-7.

26. Zaltman C, Souza HS, Castro ME, et al. Upper gastrointestinal bleeding in a Brazilian hospital: a retrospective study of endoscopic records. Arq Gastroenterol 2002; 39: 74-80.

27. Djuranovic S, Spuran M, Mijalkovic N, et al. Acute upper gastrointestinal nonvariceal bleeding--how to determine low risk patients for rebleeding and mortality after endoscopic sclerotherapy? Acta Chir Iugosl 2007; 54: 107-14. (Abstract).

28. Téllez-Avila FI, Chávez-Tapia NC, Franco-Guzmán AM, et al. Endoscopic treatment of high-risk bleeding ulcers: success, rebleeding and mortality. Rev Invest Clin 2007; 59: 419-23.