0016-5107/91/3702-0161$03.00 GASTROINTESTINAL ENDOSCOPY Copyright © 1991 by the American Society for Gastrointestinal Endoscopy
Endoscopic injection sclerotherapy versus conservative treatment for patients with unresectable hepatocellular carcinoma and bleeding esophageal varices Gin-Ho Lo, MD, Ching-Yih Lin, Kwok-Hung Lai, MD, Usman Malik, Wai-Wah Ng, MD, Fa-Yauh Lee, Shou-Dong Lee, MD, Yang-Te Tsai, Kwang-Juei Lo,
MD MD MD MD MD
Taiwan, Republic of China
We performed endoscopic injection sclerotherapy (EIS) in the treatment of 37 patients with bleeding esophageal varices due to unresectable hepatocellular carcinoma (HCC). The results were compared with those in another 33 HCC patients treated only conservatively, without EIS, during the same period. A majority of both groups died within 3 weeks after treatment. Comparing the two groups, there was no significant difference in fatal bleeding (66% vs. 75%), but significantly fewer of the EIS patients died of the index hemorrhage (43% vs. 83%; p < 0.01). Also, in the absence of portal vein thrombosis, EIS significantly reduced the risk of fatal bleeding (31% vs. 73%; p < 0.25). The mean days of survival were 32 ± 15 (range, 2 to 320) in the EIS group and 10 ± 14 (range, 2 to 270) in the compared group (p < 0.001). We conclude that EIS provides temporary control of acute esophageal variceal bleeding in patients with unresectable HCC. The major factors contributing to EIS failure are the lethal propensity of the underlying disease and portal vein thrombosis. (Gastrointest Endosc 1991;37:161-164)
Hemorrhage from esophageal varices is universally recognized as a serious, often fatal complication of portal hypertension. 1 Endoscopic injection sclerotherapy (EIS) is widely used in the management of esophageal variceal bleeding. 2-4 Several randomized, controlled trials have shown that EIS can arrest acute bleeding, decrease the frequency of rebleeding, and probably increase survival in patients with bleeding esophageal varices. 5-11 However, in order to realize long-term benefit, regularly repeated EIS is needed to achieve variceal obliteration. 1O- 12 Hemorrhage from esophageal varices in patients with unresectable hepatocellular carcinoma (HCC) is usually a terminal Received June 25, 1990. For revision September 17, 1990. Accepted November 5, 1990. Division of Gastroenterology, Department of Medicine, Veterans General Hospital, Taipei, Taiwan, Republic of China. Reprint requests: Kwok-Hung Lai, MD, Division of Gastroenterology, Department of Medicine, Veterans General Hospital, Kaohsiung, 386 TaChung 1st Rd, Kaohsiung 807, Taiwan, Republic of China. VOLUME 37, NO.2, 1991
event.13, 14 Repeated EIS can prevent rebleeding from esophageal varices,5-7,15 but it is still unknown whether an elective series of EIS is justified in those patients whose expected life span is brief. To determine whether EIS, as a supplement to otherwise supportive therapy, can provide benefit to patients with HCC complicated by bleeding varices, we performed EIS in 37 patients with this combined problem and compared the results with a similar group of 33 patients treated conservatively. PATIENTS AND METHODS
From July 1987 to February 1990, 85 consecutive patients with unresectable HCC complicated by esophageal variceal bleeding were admitted to the Taipei Veterans General Hospital. Excluded from the study were 15 patients who died within 24 hours of admission. The remaining 70 patients were enrolled in two groups: (1) 33 patients receiving standard supportive therapy (blood volume expansion, intravenous vasopressin, and balloon tamponade, as needed) 161
and (2) 37 patients receiving standard therapy supplemented by EIS. The intent of the study was fully explained to all patients and their relatives; the decision to use EIS depended on a mutually agreed upon choice, and each patient gave informed consent. Bleeding from esophageal varices was defined by the endoscopic observation of actively bleeding varices or by the presence of prominent varices in a bleeding patient when no other source of hemorrhage was evident. 14 , 15 The diagnosis of HCC was confirmed by biopsy in 28 patients and in the remainder by typical ultrasonographic, computed tomographic, or angiographic findings, associated with elevated serum a-fetoprotein levels (>400 ng/ml). Diagnosis of portal vein thrombosis was based on appropriate imaging procedures. In 35 patients, a diagnosis of HCC had been made previously, 12 days to 24 months before esophageal bleeding had occurred. In all of the patients in this series, the liver tumors were deemed unresectable because of unfavorable location or severely impaired liver function. The status of each patient was assessed at the time of presentation according to Pugh's modification of Child's classification. 16 In the EIS group, sclerotherapy was begun about 6 hours after the patients had been hemodynamically stable. An Olympus XK-I0 endoscope was used, with an inflatable balloon attached at its distal end. 17 The balloon was used to compress actively bleeding varices, thus facilitating EIS. The sclerosing agent was 5% ethanolamine or 1.5% sodium tetradecyl sulfate. Injection was within the targeted varix, as close to the esophagogastric junction as possible. A total of 8 to 20 ml of sclerosant was used at each session. The intervals between sessions depended on patient compliance and the extent of induced esophageal mucosal ulceration. Statistical analysis employed the chi-square test with Yate's correction and Student's t test.
RESULTS
There were no significant differences in clinical and laboratory features between the EIS and the control groups at entry into the study (Table 1). A previous Table 1. Patient characteristics on entry to trial
Age (yr)a Sex ratio (M/F) HBsAg (+) ALT (IU/dl)a AST (IU/dl)a Alk-P (IU/dl)a.b AFP (ng/ml)" >400 Child-Pugh class B C Main portal vein invasion Prior vasopressin infusion Prior esophageal balloon tamponade a
b
EIS group (N = 37)
Control group (N = 33)
58.2 ± 12.9 34/3 30 (81%) 120 ± 105 202 ± 185 220 ± 112
60.1 ± 9.5 31/2 28 (85%) 112 ± 65 208 ± 124 211 ± 102
25 (67%)
23 (70%)
20 (54%) 17(46%) 19(51%) 30 (81%) 22 (62%)
17(52%) 16 (48%) 18 (55%) 24 (73%) 18 (53%)
Mean ± SD. Alk-p, alkaline phosphatase; AFP, a-fetoprotein.
162
Table 2. Causes of death EIS group Control group (N = 37) (N= 33)
Causes Esophageal variceal bleeding Index bleeding Rebleeding Hepatic failure Rupture of tumor Hepatorenal syndrome Sepsis Total a
21 (66%) 9 12 7 (22%) 2 (6%) 1 (3%) 1 (3%)
24 (75%) 20 4 5 (16%) 2 (6%) 0 1 (3%)
32
32
p Nsa
Endoscopic injection sclerotherapy versus conservative treatment for patients with unresectable hepatocellular carcinoma and bleeding esophageal varices Gin-Ho Lo, MD, Ching-Yih Lin, Kwok-Hung Lai, MD, Usman Malik, Wai-Wah Ng, MD, Fa-Yauh Lee, Shou-Dong Lee, MD, Yang-Te Tsai, Kwang-Juei Lo,
MD MD MD MD MD
Taiwan, Republic of China
We performed endoscopic injection sclerotherapy (EIS) in the treatment of 37 patients with bleeding esophageal varices due to unresectable hepatocellular carcinoma (HCC). The results were compared with those in another 33 HCC patients treated only conservatively, without EIS, during the same period. A majority of both groups died within 3 weeks after treatment. Comparing the two groups, there was no significant difference in fatal bleeding (66% vs. 75%), but significantly fewer of the EIS patients died of the index hemorrhage (43% vs. 83%; p < 0.01). Also, in the absence of portal vein thrombosis, EIS significantly reduced the risk of fatal bleeding (31% vs. 73%; p < 0.25). The mean days of survival were 32 ± 15 (range, 2 to 320) in the EIS group and 10 ± 14 (range, 2 to 270) in the compared group (p < 0.001). We conclude that EIS provides temporary control of acute esophageal variceal bleeding in patients with unresectable HCC. The major factors contributing to EIS failure are the lethal propensity of the underlying disease and portal vein thrombosis. (Gastrointest Endosc 1991;37:161-164)
Hemorrhage from esophageal varices is universally recognized as a serious, often fatal complication of portal hypertension. 1 Endoscopic injection sclerotherapy (EIS) is widely used in the management of esophageal variceal bleeding. 2-4 Several randomized, controlled trials have shown that EIS can arrest acute bleeding, decrease the frequency of rebleeding, and probably increase survival in patients with bleeding esophageal varices. 5-11 However, in order to realize long-term benefit, regularly repeated EIS is needed to achieve variceal obliteration. 1O- 12 Hemorrhage from esophageal varices in patients with unresectable hepatocellular carcinoma (HCC) is usually a terminal Received June 25, 1990. For revision September 17, 1990. Accepted November 5, 1990. Division of Gastroenterology, Department of Medicine, Veterans General Hospital, Taipei, Taiwan, Republic of China. Reprint requests: Kwok-Hung Lai, MD, Division of Gastroenterology, Department of Medicine, Veterans General Hospital, Kaohsiung, 386 TaChung 1st Rd, Kaohsiung 807, Taiwan, Republic of China. VOLUME 37, NO.2, 1991
event.13, 14 Repeated EIS can prevent rebleeding from esophageal varices,5-7,15 but it is still unknown whether an elective series of EIS is justified in those patients whose expected life span is brief. To determine whether EIS, as a supplement to otherwise supportive therapy, can provide benefit to patients with HCC complicated by bleeding varices, we performed EIS in 37 patients with this combined problem and compared the results with a similar group of 33 patients treated conservatively. PATIENTS AND METHODS
From July 1987 to February 1990, 85 consecutive patients with unresectable HCC complicated by esophageal variceal bleeding were admitted to the Taipei Veterans General Hospital. Excluded from the study were 15 patients who died within 24 hours of admission. The remaining 70 patients were enrolled in two groups: (1) 33 patients receiving standard supportive therapy (blood volume expansion, intravenous vasopressin, and balloon tamponade, as needed) 161
and (2) 37 patients receiving standard therapy supplemented by EIS. The intent of the study was fully explained to all patients and their relatives; the decision to use EIS depended on a mutually agreed upon choice, and each patient gave informed consent. Bleeding from esophageal varices was defined by the endoscopic observation of actively bleeding varices or by the presence of prominent varices in a bleeding patient when no other source of hemorrhage was evident. 14 , 15 The diagnosis of HCC was confirmed by biopsy in 28 patients and in the remainder by typical ultrasonographic, computed tomographic, or angiographic findings, associated with elevated serum a-fetoprotein levels (>400 ng/ml). Diagnosis of portal vein thrombosis was based on appropriate imaging procedures. In 35 patients, a diagnosis of HCC had been made previously, 12 days to 24 months before esophageal bleeding had occurred. In all of the patients in this series, the liver tumors were deemed unresectable because of unfavorable location or severely impaired liver function. The status of each patient was assessed at the time of presentation according to Pugh's modification of Child's classification. 16 In the EIS group, sclerotherapy was begun about 6 hours after the patients had been hemodynamically stable. An Olympus XK-I0 endoscope was used, with an inflatable balloon attached at its distal end. 17 The balloon was used to compress actively bleeding varices, thus facilitating EIS. The sclerosing agent was 5% ethanolamine or 1.5% sodium tetradecyl sulfate. Injection was within the targeted varix, as close to the esophagogastric junction as possible. A total of 8 to 20 ml of sclerosant was used at each session. The intervals between sessions depended on patient compliance and the extent of induced esophageal mucosal ulceration. Statistical analysis employed the chi-square test with Yate's correction and Student's t test.
RESULTS
There were no significant differences in clinical and laboratory features between the EIS and the control groups at entry into the study (Table 1). A previous Table 1. Patient characteristics on entry to trial
Age (yr)a Sex ratio (M/F) HBsAg (+) ALT (IU/dl)a AST (IU/dl)a Alk-P (IU/dl)a.b AFP (ng/ml)" >400 Child-Pugh class B C Main portal vein invasion Prior vasopressin infusion Prior esophageal balloon tamponade a
b
EIS group (N = 37)
Control group (N = 33)
58.2 ± 12.9 34/3 30 (81%) 120 ± 105 202 ± 185 220 ± 112
60.1 ± 9.5 31/2 28 (85%) 112 ± 65 208 ± 124 211 ± 102
25 (67%)
23 (70%)
20 (54%) 17(46%) 19(51%) 30 (81%) 22 (62%)
17(52%) 16 (48%) 18 (55%) 24 (73%) 18 (53%)
Mean ± SD. Alk-p, alkaline phosphatase; AFP, a-fetoprotein.
162
Table 2. Causes of death EIS group Control group (N = 37) (N= 33)
Causes Esophageal variceal bleeding Index bleeding Rebleeding Hepatic failure Rupture of tumor Hepatorenal syndrome Sepsis Total a
21 (66%) 9 12 7 (22%) 2 (6%) 1 (3%) 1 (3%)
24 (75%) 20 4 5 (16%) 2 (6%) 0 1 (3%)
32
32
p Nsa
