- Email: [email protected]
Endoscopic hemoclip placement and epinephrine injection for Mallory-Weiss syndrome with active bleeding
Endoscopic hemoclip placement and epinephrine injection for Mallory-Weiss syndrome with active bleeding Shih-Pei Huang, MD, Hsiu-Po Wang, MD, Yi-Chia Lee, MD, Chun-Che Lin, MD, Chang-Shiu Yang, MD, Ming-Shiang Wu, MD, PhD, Jaw-Town Lin, MD, PhD Taipei, Taiwan
Background: Mallory-Weiss syndrome with active bleeding requires effective hemostasis. This is an investigation of the respective efficacy and safety of endoscopic hemoclip placement and endoscopic epinephrine injection in Mallory-Weiss syndrome. Methods: Thirty-five patients with Mallory-Weiss syndrome with spurting vessels or oozing in a university hospital were enrolled prospectively and randomly assigned to endoscopic hemoclip placement (18 patients) or endoscopic epinephrine injection (17 patients) performed by 4 endoscopists with similar clinical experiences. Demographic characteristics, endoscopic variables, and outcome parameters as well as rates of hemostasis and recurrent bleeding were analyzed. Results: The mean (SD) number of hemoclips applied was 2.5 (1.2) and the mean volume of injection was 7.9 (4.3) mL. Primary hemostasis was achieved in all 35 patients. In each group there was 1 case of recurrent bleeding. Secondary hemostasis was achieved by repeating the same procedures as at randomization in both cases. There were no significant differences in age, gender, prior ingestion of alcohol, presenting symptoms, hemoglobin level, shock, comorbid diseases, bleeding stigmata, tear location, blood transfusion, or hospitalization between the groups. There were no procedure-related complications in either group; surgery was not required in any patient. For both groups, there were no second episodes of recurrent bleeding, procedure-related complication, or need of operation. Conclusion: Endoscopic hemoclip placement and endoscopic epinephrine injection are equally effective and safe for the management of active bleeding in Mallory-Weiss syndrome, even in patients with shock or comorbid diseases. (Gastrointest Endosc 2002;55:842-6.)
Mallory-Weiss syndrome (MWS), vomitinginduced mucosal lacerations at the esophagogastric junction, remains a common cause of upper GI bleeding.1 Although the majority of patients with MWS benefit from conservative treatment, some patients, especially those with stigmata of active bleeding, unstable vital signs, and/or comorbid diseases, may require emergency treatment. Surgery,1 balloon tamponade,2 intra-arterial vasopressin infusion,3 and transcatheter embolization3 have been used to treat patients with MWS and active bleeding. With the development of therapeutic endoscopy, several endoscopic modalities, including electrocoagulation,4 ethanol injection,5 and ligation,6 have Received June 1, 2001. For revision July 24, 2001. Accepted March 1, 2002. Current affiliations: Departments of Internal Medicine and Emergency Medicine, College of Medicine, National Taiwan University Taipei, Taiwan, the Department of Internal Medicine, Chung Shan Medical and Dental College Hospital, Taichung, Taiwan, and the Department of Internal Medicine, En Chu Kong Hospital, Taipei, Taiwan. Reprint requests: Jaw-Town Lin, MD, PhD, Department of Internal Medicine, National Taiwan University Hospital, 7, Chung-Shan South Road Taipei, Taiwan. Copyright © 2002 by the American Society for Gastrointestinal Endoscopy 0016-5107/2002/$35.00 + 0 37/1/124560 doi:10.1067/mge.2002.124560 842
GASTROINTESTINAL ENDOSCOPY
been used to control upper GI bleeding. Endoscopic injection of various agents has provided effective control of bleeding from various lesions. Endoscopic injection of epinephrine and sclerotherapy was reported as effective for achieving hemostasis in bleeding MWS.7 A recent retrospective study of patients with active bleeding caused by MWS found a similar high initial hemostasis rate in a group treated by endoscopic injection of isotonic salineepinephrine solution compared with a conservatively treated group.8 Endoscopic hemoclip placement (EHP), another therapeutic alternative, has been used to stop bleeding ranging from spurting vessels to oozing lesions.9 Except for sporadic case reports, the efficacy of EHP in the treatment of bleeding MWS has not been investigated.9 This prospective randomized study was therefore conducted to compare the efficacy and safety of EHP and endoscopic epinephrine injection (EEI) in the treatment of spurting or bleeding vessels associated with MWS tears. PATIENTS AND METHODS Between October 1999 and February 2001, all consecutive patients with endoscopically verified MWS (defined as a mucosal tear near the esophagogastric junction with active bleeding, either spurting or oozing) at a single uniVOLUME 55, NO. 7, 2002
Mallory-Weiss syndrome: hemoclip placement, epinephrine injection with bleeding
A
A
B
B
S-P Huang, H-P Wang, Y-C Lee, et al.
Figure 1. A, Endoscopic view of Mallory-Weiss tear at esophagogastric junction with active bleeding in 34-year-old man with vomiting followed by hematemesis. B, View after hemoclip application to bleeding vessel (arrow); bleeding has stopped.
Figure 2. A, Endoscopic view of spurting vessel (arrow) in Mallory-Weiss tear masked by fresh blood in cardia of a 71year-old man who presented with hematemesis after vomiting. B, View after injection of epinephrine showing cessation of bleeding and a visible vessel within the tear.
versity hospital were considered for entry in this study. The patients were hospitalized and underwent endoscopy within 24 hours of the onset of bleeding. Patients were excluded if they were unable or unwilling to give informed consent for endoscopic therapy or if they were not actively bleeding at endoscopy. The study protocol was approved by the institutional review board of our hospital and informed consent was obtained from all patients enrolled. The clinical characteristics, including age, gender, presenting symptoms, prior consumption of alcohol, blood pressure, hemoglobin level, and comorbid diseases, were recorded. Shock was defined as a systolic blood pressure of less than 90 mm Hg with symptoms or signs of organ hypoperfusion. A total of 35 patients were randomized to undergo EHP or EEI by concealed allocation determined according to a random-number table. Four endoscopists (H-P.W., C-C.L., C-S.Y., S-P.H.) who had the same training course in therapeutic endoscopy at our university hospital performed the diagnostic and therapeutic endoscopic procedures. They did
not participate in the postprocedure care of the patients, which was conducted by other physicians. Endoscopy was performed with a standard upper endoscope (GJF-230, Olympus, Optical, Co., Ltd., Tokyo, Japan). The location of the tear, number of bleeding points, and type of stigmata were recorded based on endoscopic findings. In the EHP group, hemoclips (MDS50, Olympus) were applied to bleeding points with a clip application device (HX-3L, Olympus) (Fig. 1). In the EEI group, 1 to 3 mL of a 1:10,000 solution of epinephrine (prepared by mixing 1 mL of 1:1000 epinephrine and 9 mL 0.9% saline solution) were injected into the surrounding area close to bleeders by an injection needle through the accessory channel (Fig. 2). In the EEI group, shots of 1 to 3 mL 1:10,000 epinephrine, prepared by mixing 1 mL 1:1000 epinephrine and 9 mL 0.9% saline solution, were injected into the area surrounding and close to the bleeding points with an injection needle. After therapy in both groups, the bleeding points were gently irrigated with 0.9% saline solution to evaluate the hemostatic effect.
VOLUME 55, NO. 7, 2002
GASTROINTESTINAL ENDOSCOPY
843
S-P Huang, H-P Wang, Y-C Lee, et al.
Mallory-Weiss syndrome: hemoclip placement, epinephrine injection with bleeding
Table 1. Clinical and endoscopic characteristics of EHP and EEI groups
Age (y) Gender (male/female) Alcohol ingestion before vomiting Hematemesis/melena Shock Hemoglobin level (gm/dL) Comorbid disease Bleeder number Bleeding stigmata Spurting vessel Oozing vessel Tear location Distal esophagus Esophagogastric junction Cardia
EHP (n = 18)
EEI (n = 17)
p Value
55.3 (18.6) 15/3 5 (28%)
50.6 (17.9) 14/3 0
0.45 1.00 0.45
18/0 6 (33%) 10.4 (2.9) 11 (61%) 1.3 (0.6)
16/1 3 (18%) 10.4 (6.2) 9 (53%) 1.3 (0.7)
0.49 0.44 0.99 0.74 0.94 0.09
6 (33%) 12 (67%)
1 (6%) 16 (94%)
1 (6%) 15 (83%) 2 (11%)
2 (12%) 12 (71%) 3 (17%)
0.60
Quantitative data are expressed as mean (SD).
Primary hemostasis was defined as endoscopically verified cessation of bleeding for at least 1 minute after hemoclip application or injection during the first endoscopic session. Patients who continued to bleed despite EHP or EEI with refractory shock despite blood transfusion and intravenous fluid challenge underwent emergency surgery. A first episode of recurrent bleeding was defined as a recurrence of hematemesis or bloody nasogastric aspirate, or a decrease in hemoglobin concentration of more than 1.5 gm/dL within 24 hours after successfully achieving primary hemostasis. Patients with a first episode of recurrent bleeding underwent a second endoscopy and were treated with the same therapeutic modality as at randomization. Secondary hemostasis was defined as cessation of first episode of recurrent bleeding after a second session of therapeutic endoscopy. A second episode of recurrent bleeding was defined as a recurrence of hematemesis or bloody nasogastric aspirate, or a decrease in hemoglobin concentration of more than 1.5 gm/dL within 24 hours of a second therapeutic procedure. Patients with a second episode of recurrent bleeding underwent emergency surgery. Crossover was not permitted between the EHP and EEI treatment groups. The long-term outcome of the treatments was evaluated by staff who were blinded to the type of therapy delivered and who followed the patients for 30 days including the initial period of hospitalization. Permanent hemostasis was defined as the absence of recurrent bleeding during this period. The efficacy of the 2 hemostatic procedures was evaluated according to rates of primary and secondary hemostasis, first and second episodes of recurrent bleeding and permanent hemostasis, units of blood transfused, need for emergency operation, length of hospital stay, and hospital mortality rate. The primary outcomes were rates of hemostasis and recurrent bleeding. Quantitative data were summarized as mean (SD) and compared with the Mann-Whitney U test. Count data were analyzed by the Fisher exact test. A p value less than 844
GASTROINTESTINAL ENDOSCOPY
Table 2. Efficacy of endoscopic hemostasis and clinical outcome of EHP and EEI groups
Primary outcomes Primary hemostasis First episode of recurrent bleeding Permanent primary hemostasis Secondary hemostasis Other outcomes Blood transfusion (mean units) Hospital death Hospital stay (mean no. of days)
EHP (n = 18)
EEI (n = 17)
p Value
18 (100%) 1 (6%)
17 (100%) 1 (6%)
1.00 1.00
17 (94%)
16 (94%)
1.00
1 (100%)
1 (100%)
1.00
2.2 (3.2)
1.9 (1.7)
0.75
1 (6%) 8.2 (9.9)
0 4.5 (6.9)
1.00 0.21
Quantitative data are expressed as mean (SD).
0.05 was considered significant. The analysis was performed with statistical software package (SPSS 10.0 version for Windows, SPSS, Chicago, Ill.).
RESULTS A diagnosis of MWS was made in 102 patients during the study period, of which 35 patients (34%) met the entry criteria and were enrolled. Among them, 18 were randomized to EHP and 17 to EEI (Table 1). There were no significant differences between these groups with respect to age, gender, hemoglobin level, number of bleeding points, stigmata of bleeding, or locations of the lacerations. There was a predominance of men in both, and hematemesis was the presenting symptom in all but 1 patient with melena in the EEI group. Five patients (28%) in the EHP group had a history of alcohol consumption before the episodes of MWS; there was no history of alcohol ingestion for any patient in the EEI group (p = 0.45). Before endoscopy, 6 patients (33%) in the EHP group and 3 (18%) of the EEI group were in shock (p = 0.44). Comorbid diseases were noted in 11 patients (61%) in the EHP group and 9 (53%) of the EEI group (p = 0.74). The comorbid diseases in the EHP group included acute myocardial infarction, stroke, nasopharyngeal carcinoma, sepsis, lymphoma, uremia, pneumonia, respiratory failure, and ileus with jejunal bleeding; those in the EEI group included pancreatitis, uremia, systemic lupus erythematosis, hepatocellular carcinoma, choledocholithiasis, chronic obstructive pulmonary disease, and coronary artery disease. There was no significant difference between the groups with respect to frequency of a spurting vessel and oozing, although in the EHP group there were more spurting vessels (p = 0.09). The majority of the lacerations were at the esophagogastric junction in both groups (p = 0.60). VOLUME 55, NO. 7, 2002
Mallory-Weiss syndrome: hemoclip placement, epinephrine injection with bleeding
The efficacy of the 2 hemostatic procedures and clinical outcomes for each group are summarized in Table 2. Primary hemostasis was achieved with both hemostatic modalities in all patients. The mean number of hemoclips applied was 2.5 (1.2) and the mean volume of the epinephrine solution injected was 7.9 (4.3) mL. One patient in each group had recurrent bleeding. Secondary hemostasis was achieved in both patients by using the same treatment modality as at randomization. There were no differences in primary outcomes including rates of primary hemostasis, recurrent bleeding, and permanent primary hemostasis. No patient had more than 1 episode of recurrent bleeding and none underwent surgery. The number of blood transfusions was 2.2 (3.2) units in the EHP group and 1.9 (1.7) units in the EEI group (p = 0.75). The mean length of hospital stay was longer in the EHP (8.2 days) compared with the EEI group (4.5 days), but the difference was not significant (p = 0.21). After a long hospital course (44 days), 1 patient in the EHP group died, without further bleeding, because of a complicated acute myocardial infarction. DISCUSSION MWS accounts for 5% to 15% of cases of upper GI bleeding.1 The majority of patients respond to conservative treatment. However, when the bleeding is active, it persists in 50% of cases.10 Therefore, intervention to achieve hemostasis is mandatory in patients at high risk for persistent bleeding.4 In our series, 34% of patients with MWS had active bleeding, either spurting or oozing. This frequency is within the range of 25% to 42% reported.11 There were no significant differences between the 2 treatment groups with regard to age, gender, presenting symptoms, prior alcohol ingestion, frequency of comorbid diseases, hemodynamic status, or hemoglobin levels. The predominancy of men among patients with MWS and the tendency for the laceration to be located at the esophagogastric junction have been noted previously.12,13 Endoscopic methods are generally accepted as effective for the treatment of acute GI bleeding. The retrospective study of Peng et al.8 found equally high rates of hemostasis in patients with active bleeding caused by MWS treated by endoscopic injection of isotonic saline-epinephrine solution versus conservatively treated patients (respectively, 93% and 95%). However, the initial hematocrit in the injection group was significantly lower than that in the conservative treatment group, which may indicate more severe bleeding in the former group. In another study, endoscopic sclerotherapy with 1% polidocanol and epinephrine was found effective in the treatment of 13 patients with MWS and active bleeding VOLUME 55, NO. 7, 2002
S-P Huang, H-P Wang, Y-C Lee, et al.
or a visible vessel.7 In the present study, a 0.9% saline solution of epinephrine was used in the EEI group rather than a combination of epinephrinepolidocanol or epinephrine-hypertonic saline solution. The rate of primary hemostasis was 100% in patients with MWS and active bleeding treated by EEI. EHP was introduced by Hayashi et al.9 in 1975 and has been shown to be effective in the control of GI bleeding from a wide variety of sources.8 There were, however, only sporadic case reports of the efficacy of EHP for active bleeding caused by MWS.9,14 A study comparing the hemostatic efficacy of EHP with injection of hypertonic saline-epinephrine solution injection in bleeding peptic ulcers found that EHP is more effective and safer, and that combined therapy did not have an advantage over EHP alone.15 In the present study, the success rate for EHP in the primary control of bleeding in MWS was 100%. It has been proposed that hemodynamically stable patients with MWS and no major diseases or active bleeding can be managed with a brief period of observation.16 In the present study, shock and comorbid diseases were present in, respectively, 6 (33%) and 11 patients (61%) in the EHP group, and, respectively, in 3 (18%) and 9 patients (53%) in the EEI. Despite this, primary hemostasis was achieved by both EHP and EEI in all patients. Furthermore, EHP and EEI both achieved the same low rate of recurrent bleeding (6%) and high rates of permanent primary hemostasis. The single patient with recurrent bleeding in the EHP group had a spurting vessel and shock but no comorbid disease. The only patient in whom bleeding recurred in the EEI group had an oozing tear and uremia but not shock. The recurrent bleeding in these 2 patients was successfully controlled by using the same modality as at randomization. There was no procedure-related complication in either group. The results of the present study indicate that both EHP and EEI are safe and effective for treatment of patients with bleeding caused by MWS including patients at high risk of recurrent bleeding. In addition, should bleeding recur, both methods are equally effective for achieving secondary hemostasis. In a review of 69 cases in 1981, Hastings et al.17 found that patients with MWS and persistent bleeding after transfusion of more than 1.5 L of blood required surgery. The number of blood transfusions was 2.2 (3.2) units in the EHP group and 1.9 (1.7) units in the EEI group (p = 0.82) in the current study. No patient in either group required surgery. It appears, therefore, that both EHP and EEI markedly reduce the need for operation in patients with bleeding MWS. One patient died in the EHP group after a prolonged hospitalization, death being caused by a complicated myocardial infarction GASTROINTESTINAL ENDOSCOPY
845
S-P Huang, H-P Wang, Y-C Lee, et al.
Mallory-Weiss syndrome: hemoclip placement, epinephrine injection with bleeding
rather than bleeding, suggesting that the morbidity and mortality associated with bleeding MWS are influenced by the presence of other illnesses. In summary, EHP and EEI are equally safe and effective for control of bleeding, spurting, or oozing in patients with MWS, including those who are hemodynamically unstable and/or have major comorbid illness. Moreover, after treatment the frequency of recurrent bleeding is low and the rate of permanent hemostasis is high. REFERENCES 1. Sugawa C, Benishek D, Walt AS. Mallory-Weiss syndrome. A study of 224 patients. Am J Surg 1983;145:30-3. 2. Welch GH, McArdle CS, Anderson JR. Balloon tamponade for the control of Mallory-Weiss haemorrhage in patients with coagulation defects. Br J Surg 1987;74:610-1. 3. Fisher RG, Schwartz JT, Graham DY. Angiotherapy with Mallory-Weiss tear. AJR Am J Roentgenol 1980;134:679-84. 4. Laine L. Multipolar electrocoagulation in the treatment of active upper gastrointestinal tract hemorrhage. New Engl J Med 1987;316:1613-7. 5. Sugawa C, Fujita Y, Ikeda T, Walt AJ. Endoscopic hemostasis of bleeding of the upper gastrointestinal tract by local injection of ninety-eight per cent dehydrated ethanol. Surg Gynecol Obstet 1986;162:159-63. 6. Wong RM, Ota S, Katoh A, Yamauchi A, Arai K, Kaneko K, et al. Endoscopic ligation for non-esophageal variceal upper gastrointestinal hemorrhage. Endoscopy 1998;30:774-7. 7. Bataller R, Llach J, Salineron JM, Elizalde JI, Mas A, Pique TM, et al. Endoscopic sclerotherapy in upper gastrointestinal bleeding due to the Mallory-Weiss syndrome. Am J Gastroenterol 1994;89:2147-50.
846
GASTROINTESTINAL ENDOSCOPY
8. Peng YC, Tung CF, Chow WK, Chang CS, Chen GH, Hu WH, et al. Efficacy of endoscopic isotonic saline-epinephrine injection for the management of active Mallory-Weiss tears. J Clin Gastroenterol 2001;32:119-22. 9. Hayashi T, Yonezawa M, Kuwabara T, Kudoh I. The study on staunch clip for the treatment by endoscopy. Gastroenterol Endosc 1975;17:92-101. 10. Jensen DM, Kovacs TOG, Machicado GA, Randall GM, Sue M. CURE Hemostasis Research Unit. Prospective study of the stigmata of hemorrhage and endoscopic or medical treatment for bleeding Mallory Weiss tears [abstract]. Gastrointest Endosc 1992;38:235. 11. Lum DF, McQuaid K, Lee JG. Endoscopic hemostasis of nonvariceal, non-peptic ulcer hemorrhage. Gastrointest Endosc Clin N Am 1997;7:657-70. 12. Bubrick Mp, Ludeen JW, Onstad GR, Hitchcock CR. MalloryWeiss syndrome: analysis of fifty-nine patients. Surgery 1980;88:400-5. 13. Knauer CM. Mallory-Weiss syndrome: characterization of 75 Mallory-Weiss lacerations in 528 patients with upper gastrointestinal hemorrhage. Gastroenterology 1976;71:5-8. 14. Ohta S, Yukioka T, Ohta S, Miyagatani Y, Matsuda H, Shimazaki S. Hemostasis with endoscopic hemoclipping for severe gastrointestinal bleeding in critical ill patients. Am J Gastroenterol 1996;91:701-4. 15. Chung IK, Ham JS, Kim HS, Park HS, Lee MH, Kim SJ. Comparison of the hemostatic efficacy of the endoscopic hemoclip method with hypertonic saline-epinephrine injection and a combination of the two for the management of bleeding peptic ulcers. Gastrointest Endosc 1999;49:13-8. 16. Bharucha AE, Gostout CJ, Balm RK. Clinical and endoscopic risk factors in the Mallory-Weiss syndrome. Am J Gastroenterol 1997;92:805-8. 17. Hastings PR, Peters KW, Cohn I. Mallory-Weiss syndrome. Review of 69 cases. Am J Surg 1981;142:560-2.
VOLUME 55, NO. 7, 2002
Background: Mallory-Weiss syndrome with active bleeding requires effective hemostasis. This is an investigation of the respective efficacy and safety of endoscopic hemoclip placement and endoscopic epinephrine injection in Mallory-Weiss syndrome. Methods: Thirty-five patients with Mallory-Weiss syndrome with spurting vessels or oozing in a university hospital were enrolled prospectively and randomly assigned to endoscopic hemoclip placement (18 patients) or endoscopic epinephrine injection (17 patients) performed by 4 endoscopists with similar clinical experiences. Demographic characteristics, endoscopic variables, and outcome parameters as well as rates of hemostasis and recurrent bleeding were analyzed. Results: The mean (SD) number of hemoclips applied was 2.5 (1.2) and the mean volume of injection was 7.9 (4.3) mL. Primary hemostasis was achieved in all 35 patients. In each group there was 1 case of recurrent bleeding. Secondary hemostasis was achieved by repeating the same procedures as at randomization in both cases. There were no significant differences in age, gender, prior ingestion of alcohol, presenting symptoms, hemoglobin level, shock, comorbid diseases, bleeding stigmata, tear location, blood transfusion, or hospitalization between the groups. There were no procedure-related complications in either group; surgery was not required in any patient. For both groups, there were no second episodes of recurrent bleeding, procedure-related complication, or need of operation. Conclusion: Endoscopic hemoclip placement and endoscopic epinephrine injection are equally effective and safe for the management of active bleeding in Mallory-Weiss syndrome, even in patients with shock or comorbid diseases. (Gastrointest Endosc 2002;55:842-6.)
Mallory-Weiss syndrome (MWS), vomitinginduced mucosal lacerations at the esophagogastric junction, remains a common cause of upper GI bleeding.1 Although the majority of patients with MWS benefit from conservative treatment, some patients, especially those with stigmata of active bleeding, unstable vital signs, and/or comorbid diseases, may require emergency treatment. Surgery,1 balloon tamponade,2 intra-arterial vasopressin infusion,3 and transcatheter embolization3 have been used to treat patients with MWS and active bleeding. With the development of therapeutic endoscopy, several endoscopic modalities, including electrocoagulation,4 ethanol injection,5 and ligation,6 have Received June 1, 2001. For revision July 24, 2001. Accepted March 1, 2002. Current affiliations: Departments of Internal Medicine and Emergency Medicine, College of Medicine, National Taiwan University Taipei, Taiwan, the Department of Internal Medicine, Chung Shan Medical and Dental College Hospital, Taichung, Taiwan, and the Department of Internal Medicine, En Chu Kong Hospital, Taipei, Taiwan. Reprint requests: Jaw-Town Lin, MD, PhD, Department of Internal Medicine, National Taiwan University Hospital, 7, Chung-Shan South Road Taipei, Taiwan. Copyright © 2002 by the American Society for Gastrointestinal Endoscopy 0016-5107/2002/$35.00 + 0 37/1/124560 doi:10.1067/mge.2002.124560 842
GASTROINTESTINAL ENDOSCOPY
been used to control upper GI bleeding. Endoscopic injection of various agents has provided effective control of bleeding from various lesions. Endoscopic injection of epinephrine and sclerotherapy was reported as effective for achieving hemostasis in bleeding MWS.7 A recent retrospective study of patients with active bleeding caused by MWS found a similar high initial hemostasis rate in a group treated by endoscopic injection of isotonic salineepinephrine solution compared with a conservatively treated group.8 Endoscopic hemoclip placement (EHP), another therapeutic alternative, has been used to stop bleeding ranging from spurting vessels to oozing lesions.9 Except for sporadic case reports, the efficacy of EHP in the treatment of bleeding MWS has not been investigated.9 This prospective randomized study was therefore conducted to compare the efficacy and safety of EHP and endoscopic epinephrine injection (EEI) in the treatment of spurting or bleeding vessels associated with MWS tears. PATIENTS AND METHODS Between October 1999 and February 2001, all consecutive patients with endoscopically verified MWS (defined as a mucosal tear near the esophagogastric junction with active bleeding, either spurting or oozing) at a single uniVOLUME 55, NO. 7, 2002
Mallory-Weiss syndrome: hemoclip placement, epinephrine injection with bleeding
A
A
B
B
S-P Huang, H-P Wang, Y-C Lee, et al.
Figure 1. A, Endoscopic view of Mallory-Weiss tear at esophagogastric junction with active bleeding in 34-year-old man with vomiting followed by hematemesis. B, View after hemoclip application to bleeding vessel (arrow); bleeding has stopped.
Figure 2. A, Endoscopic view of spurting vessel (arrow) in Mallory-Weiss tear masked by fresh blood in cardia of a 71year-old man who presented with hematemesis after vomiting. B, View after injection of epinephrine showing cessation of bleeding and a visible vessel within the tear.
versity hospital were considered for entry in this study. The patients were hospitalized and underwent endoscopy within 24 hours of the onset of bleeding. Patients were excluded if they were unable or unwilling to give informed consent for endoscopic therapy or if they were not actively bleeding at endoscopy. The study protocol was approved by the institutional review board of our hospital and informed consent was obtained from all patients enrolled. The clinical characteristics, including age, gender, presenting symptoms, prior consumption of alcohol, blood pressure, hemoglobin level, and comorbid diseases, were recorded. Shock was defined as a systolic blood pressure of less than 90 mm Hg with symptoms or signs of organ hypoperfusion. A total of 35 patients were randomized to undergo EHP or EEI by concealed allocation determined according to a random-number table. Four endoscopists (H-P.W., C-C.L., C-S.Y., S-P.H.) who had the same training course in therapeutic endoscopy at our university hospital performed the diagnostic and therapeutic endoscopic procedures. They did
not participate in the postprocedure care of the patients, which was conducted by other physicians. Endoscopy was performed with a standard upper endoscope (GJF-230, Olympus, Optical, Co., Ltd., Tokyo, Japan). The location of the tear, number of bleeding points, and type of stigmata were recorded based on endoscopic findings. In the EHP group, hemoclips (MDS50, Olympus) were applied to bleeding points with a clip application device (HX-3L, Olympus) (Fig. 1). In the EEI group, 1 to 3 mL of a 1:10,000 solution of epinephrine (prepared by mixing 1 mL of 1:1000 epinephrine and 9 mL 0.9% saline solution) were injected into the surrounding area close to bleeders by an injection needle through the accessory channel (Fig. 2). In the EEI group, shots of 1 to 3 mL 1:10,000 epinephrine, prepared by mixing 1 mL 1:1000 epinephrine and 9 mL 0.9% saline solution, were injected into the area surrounding and close to the bleeding points with an injection needle. After therapy in both groups, the bleeding points were gently irrigated with 0.9% saline solution to evaluate the hemostatic effect.
VOLUME 55, NO. 7, 2002
GASTROINTESTINAL ENDOSCOPY
843
S-P Huang, H-P Wang, Y-C Lee, et al.
Mallory-Weiss syndrome: hemoclip placement, epinephrine injection with bleeding
Table 1. Clinical and endoscopic characteristics of EHP and EEI groups
Age (y) Gender (male/female) Alcohol ingestion before vomiting Hematemesis/melena Shock Hemoglobin level (gm/dL) Comorbid disease Bleeder number Bleeding stigmata Spurting vessel Oozing vessel Tear location Distal esophagus Esophagogastric junction Cardia
EHP (n = 18)
EEI (n = 17)
p Value
55.3 (18.6) 15/3 5 (28%)
50.6 (17.9) 14/3 0
0.45 1.00 0.45
18/0 6 (33%) 10.4 (2.9) 11 (61%) 1.3 (0.6)
16/1 3 (18%) 10.4 (6.2) 9 (53%) 1.3 (0.7)
0.49 0.44 0.99 0.74 0.94 0.09
6 (33%) 12 (67%)
1 (6%) 16 (94%)
1 (6%) 15 (83%) 2 (11%)
2 (12%) 12 (71%) 3 (17%)
0.60
Quantitative data are expressed as mean (SD).
Primary hemostasis was defined as endoscopically verified cessation of bleeding for at least 1 minute after hemoclip application or injection during the first endoscopic session. Patients who continued to bleed despite EHP or EEI with refractory shock despite blood transfusion and intravenous fluid challenge underwent emergency surgery. A first episode of recurrent bleeding was defined as a recurrence of hematemesis or bloody nasogastric aspirate, or a decrease in hemoglobin concentration of more than 1.5 gm/dL within 24 hours after successfully achieving primary hemostasis. Patients with a first episode of recurrent bleeding underwent a second endoscopy and were treated with the same therapeutic modality as at randomization. Secondary hemostasis was defined as cessation of first episode of recurrent bleeding after a second session of therapeutic endoscopy. A second episode of recurrent bleeding was defined as a recurrence of hematemesis or bloody nasogastric aspirate, or a decrease in hemoglobin concentration of more than 1.5 gm/dL within 24 hours of a second therapeutic procedure. Patients with a second episode of recurrent bleeding underwent emergency surgery. Crossover was not permitted between the EHP and EEI treatment groups. The long-term outcome of the treatments was evaluated by staff who were blinded to the type of therapy delivered and who followed the patients for 30 days including the initial period of hospitalization. Permanent hemostasis was defined as the absence of recurrent bleeding during this period. The efficacy of the 2 hemostatic procedures was evaluated according to rates of primary and secondary hemostasis, first and second episodes of recurrent bleeding and permanent hemostasis, units of blood transfused, need for emergency operation, length of hospital stay, and hospital mortality rate. The primary outcomes were rates of hemostasis and recurrent bleeding. Quantitative data were summarized as mean (SD) and compared with the Mann-Whitney U test. Count data were analyzed by the Fisher exact test. A p value less than 844
GASTROINTESTINAL ENDOSCOPY
Table 2. Efficacy of endoscopic hemostasis and clinical outcome of EHP and EEI groups
Primary outcomes Primary hemostasis First episode of recurrent bleeding Permanent primary hemostasis Secondary hemostasis Other outcomes Blood transfusion (mean units) Hospital death Hospital stay (mean no. of days)
EHP (n = 18)
EEI (n = 17)
p Value
18 (100%) 1 (6%)
17 (100%) 1 (6%)
1.00 1.00
17 (94%)
16 (94%)
1.00
1 (100%)
1 (100%)
1.00
2.2 (3.2)
1.9 (1.7)
0.75
1 (6%) 8.2 (9.9)
0 4.5 (6.9)
1.00 0.21
Quantitative data are expressed as mean (SD).
0.05 was considered significant. The analysis was performed with statistical software package (SPSS 10.0 version for Windows, SPSS, Chicago, Ill.).
RESULTS A diagnosis of MWS was made in 102 patients during the study period, of which 35 patients (34%) met the entry criteria and were enrolled. Among them, 18 were randomized to EHP and 17 to EEI (Table 1). There were no significant differences between these groups with respect to age, gender, hemoglobin level, number of bleeding points, stigmata of bleeding, or locations of the lacerations. There was a predominance of men in both, and hematemesis was the presenting symptom in all but 1 patient with melena in the EEI group. Five patients (28%) in the EHP group had a history of alcohol consumption before the episodes of MWS; there was no history of alcohol ingestion for any patient in the EEI group (p = 0.45). Before endoscopy, 6 patients (33%) in the EHP group and 3 (18%) of the EEI group were in shock (p = 0.44). Comorbid diseases were noted in 11 patients (61%) in the EHP group and 9 (53%) of the EEI group (p = 0.74). The comorbid diseases in the EHP group included acute myocardial infarction, stroke, nasopharyngeal carcinoma, sepsis, lymphoma, uremia, pneumonia, respiratory failure, and ileus with jejunal bleeding; those in the EEI group included pancreatitis, uremia, systemic lupus erythematosis, hepatocellular carcinoma, choledocholithiasis, chronic obstructive pulmonary disease, and coronary artery disease. There was no significant difference between the groups with respect to frequency of a spurting vessel and oozing, although in the EHP group there were more spurting vessels (p = 0.09). The majority of the lacerations were at the esophagogastric junction in both groups (p = 0.60). VOLUME 55, NO. 7, 2002
Mallory-Weiss syndrome: hemoclip placement, epinephrine injection with bleeding
The efficacy of the 2 hemostatic procedures and clinical outcomes for each group are summarized in Table 2. Primary hemostasis was achieved with both hemostatic modalities in all patients. The mean number of hemoclips applied was 2.5 (1.2) and the mean volume of the epinephrine solution injected was 7.9 (4.3) mL. One patient in each group had recurrent bleeding. Secondary hemostasis was achieved in both patients by using the same treatment modality as at randomization. There were no differences in primary outcomes including rates of primary hemostasis, recurrent bleeding, and permanent primary hemostasis. No patient had more than 1 episode of recurrent bleeding and none underwent surgery. The number of blood transfusions was 2.2 (3.2) units in the EHP group and 1.9 (1.7) units in the EEI group (p = 0.75). The mean length of hospital stay was longer in the EHP (8.2 days) compared with the EEI group (4.5 days), but the difference was not significant (p = 0.21). After a long hospital course (44 days), 1 patient in the EHP group died, without further bleeding, because of a complicated acute myocardial infarction. DISCUSSION MWS accounts for 5% to 15% of cases of upper GI bleeding.1 The majority of patients respond to conservative treatment. However, when the bleeding is active, it persists in 50% of cases.10 Therefore, intervention to achieve hemostasis is mandatory in patients at high risk for persistent bleeding.4 In our series, 34% of patients with MWS had active bleeding, either spurting or oozing. This frequency is within the range of 25% to 42% reported.11 There were no significant differences between the 2 treatment groups with regard to age, gender, presenting symptoms, prior alcohol ingestion, frequency of comorbid diseases, hemodynamic status, or hemoglobin levels. The predominancy of men among patients with MWS and the tendency for the laceration to be located at the esophagogastric junction have been noted previously.12,13 Endoscopic methods are generally accepted as effective for the treatment of acute GI bleeding. The retrospective study of Peng et al.8 found equally high rates of hemostasis in patients with active bleeding caused by MWS treated by endoscopic injection of isotonic saline-epinephrine solution versus conservatively treated patients (respectively, 93% and 95%). However, the initial hematocrit in the injection group was significantly lower than that in the conservative treatment group, which may indicate more severe bleeding in the former group. In another study, endoscopic sclerotherapy with 1% polidocanol and epinephrine was found effective in the treatment of 13 patients with MWS and active bleeding VOLUME 55, NO. 7, 2002
S-P Huang, H-P Wang, Y-C Lee, et al.
or a visible vessel.7 In the present study, a 0.9% saline solution of epinephrine was used in the EEI group rather than a combination of epinephrinepolidocanol or epinephrine-hypertonic saline solution. The rate of primary hemostasis was 100% in patients with MWS and active bleeding treated by EEI. EHP was introduced by Hayashi et al.9 in 1975 and has been shown to be effective in the control of GI bleeding from a wide variety of sources.8 There were, however, only sporadic case reports of the efficacy of EHP for active bleeding caused by MWS.9,14 A study comparing the hemostatic efficacy of EHP with injection of hypertonic saline-epinephrine solution injection in bleeding peptic ulcers found that EHP is more effective and safer, and that combined therapy did not have an advantage over EHP alone.15 In the present study, the success rate for EHP in the primary control of bleeding in MWS was 100%. It has been proposed that hemodynamically stable patients with MWS and no major diseases or active bleeding can be managed with a brief period of observation.16 In the present study, shock and comorbid diseases were present in, respectively, 6 (33%) and 11 patients (61%) in the EHP group, and, respectively, in 3 (18%) and 9 patients (53%) in the EEI. Despite this, primary hemostasis was achieved by both EHP and EEI in all patients. Furthermore, EHP and EEI both achieved the same low rate of recurrent bleeding (6%) and high rates of permanent primary hemostasis. The single patient with recurrent bleeding in the EHP group had a spurting vessel and shock but no comorbid disease. The only patient in whom bleeding recurred in the EEI group had an oozing tear and uremia but not shock. The recurrent bleeding in these 2 patients was successfully controlled by using the same modality as at randomization. There was no procedure-related complication in either group. The results of the present study indicate that both EHP and EEI are safe and effective for treatment of patients with bleeding caused by MWS including patients at high risk of recurrent bleeding. In addition, should bleeding recur, both methods are equally effective for achieving secondary hemostasis. In a review of 69 cases in 1981, Hastings et al.17 found that patients with MWS and persistent bleeding after transfusion of more than 1.5 L of blood required surgery. The number of blood transfusions was 2.2 (3.2) units in the EHP group and 1.9 (1.7) units in the EEI group (p = 0.82) in the current study. No patient in either group required surgery. It appears, therefore, that both EHP and EEI markedly reduce the need for operation in patients with bleeding MWS. One patient died in the EHP group after a prolonged hospitalization, death being caused by a complicated myocardial infarction GASTROINTESTINAL ENDOSCOPY
845
S-P Huang, H-P Wang, Y-C Lee, et al.
Mallory-Weiss syndrome: hemoclip placement, epinephrine injection with bleeding
rather than bleeding, suggesting that the morbidity and mortality associated with bleeding MWS are influenced by the presence of other illnesses. In summary, EHP and EEI are equally safe and effective for control of bleeding, spurting, or oozing in patients with MWS, including those who are hemodynamically unstable and/or have major comorbid illness. Moreover, after treatment the frequency of recurrent bleeding is low and the rate of permanent hemostasis is high. REFERENCES 1. Sugawa C, Benishek D, Walt AS. Mallory-Weiss syndrome. A study of 224 patients. Am J Surg 1983;145:30-3. 2. Welch GH, McArdle CS, Anderson JR. Balloon tamponade for the control of Mallory-Weiss haemorrhage in patients with coagulation defects. Br J Surg 1987;74:610-1. 3. Fisher RG, Schwartz JT, Graham DY. Angiotherapy with Mallory-Weiss tear. AJR Am J Roentgenol 1980;134:679-84. 4. Laine L. Multipolar electrocoagulation in the treatment of active upper gastrointestinal tract hemorrhage. New Engl J Med 1987;316:1613-7. 5. Sugawa C, Fujita Y, Ikeda T, Walt AJ. Endoscopic hemostasis of bleeding of the upper gastrointestinal tract by local injection of ninety-eight per cent dehydrated ethanol. Surg Gynecol Obstet 1986;162:159-63. 6. Wong RM, Ota S, Katoh A, Yamauchi A, Arai K, Kaneko K, et al. Endoscopic ligation for non-esophageal variceal upper gastrointestinal hemorrhage. Endoscopy 1998;30:774-7. 7. Bataller R, Llach J, Salineron JM, Elizalde JI, Mas A, Pique TM, et al. Endoscopic sclerotherapy in upper gastrointestinal bleeding due to the Mallory-Weiss syndrome. Am J Gastroenterol 1994;89:2147-50.
846
GASTROINTESTINAL ENDOSCOPY
8. Peng YC, Tung CF, Chow WK, Chang CS, Chen GH, Hu WH, et al. Efficacy of endoscopic isotonic saline-epinephrine injection for the management of active Mallory-Weiss tears. J Clin Gastroenterol 2001;32:119-22. 9. Hayashi T, Yonezawa M, Kuwabara T, Kudoh I. The study on staunch clip for the treatment by endoscopy. Gastroenterol Endosc 1975;17:92-101. 10. Jensen DM, Kovacs TOG, Machicado GA, Randall GM, Sue M. CURE Hemostasis Research Unit. Prospective study of the stigmata of hemorrhage and endoscopic or medical treatment for bleeding Mallory Weiss tears [abstract]. Gastrointest Endosc 1992;38:235. 11. Lum DF, McQuaid K, Lee JG. Endoscopic hemostasis of nonvariceal, non-peptic ulcer hemorrhage. Gastrointest Endosc Clin N Am 1997;7:657-70. 12. Bubrick Mp, Ludeen JW, Onstad GR, Hitchcock CR. MalloryWeiss syndrome: analysis of fifty-nine patients. Surgery 1980;88:400-5. 13. Knauer CM. Mallory-Weiss syndrome: characterization of 75 Mallory-Weiss lacerations in 528 patients with upper gastrointestinal hemorrhage. Gastroenterology 1976;71:5-8. 14. Ohta S, Yukioka T, Ohta S, Miyagatani Y, Matsuda H, Shimazaki S. Hemostasis with endoscopic hemoclipping for severe gastrointestinal bleeding in critical ill patients. Am J Gastroenterol 1996;91:701-4. 15. Chung IK, Ham JS, Kim HS, Park HS, Lee MH, Kim SJ. Comparison of the hemostatic efficacy of the endoscopic hemoclip method with hypertonic saline-epinephrine injection and a combination of the two for the management of bleeding peptic ulcers. Gastrointest Endosc 1999;49:13-8. 16. Bharucha AE, Gostout CJ, Balm RK. Clinical and endoscopic risk factors in the Mallory-Weiss syndrome. Am J Gastroenterol 1997;92:805-8. 17. Hastings PR, Peters KW, Cohn I. Mallory-Weiss syndrome. Review of 69 cases. Am J Surg 1981;142:560-2.
VOLUME 55, NO. 7, 2002