World J. Surg. 16, 97-105, 1992
World Journal of Surgery O 1992 by the Socit~t6 lnternationale de Chlrur~e
Original Scientific Reports Hepatic Resection for Hepatocellular Carcinoma Masahiro Suenaga, M.D., Akimasa Nakao, M.D., Akio Harada, M.D., Toshiaki N o n a m i , M.D., Yoshikatsu Okada, M.D., H a y a t o Sugiura, M.D., Shinichi Uehara, M.D., and Hiroshi Takagi, M.D. Surgical Service, Nagoya Memorial Hospital, and Department of Surgery II, Nagoya University, School of Medicine, Nagoya, Japan One hundred and eighteen patients underwent hepatic resection for hepatocellular carcinoma from 1979 to 1987. Ninety-eight of these patients had co-existing cirrhosis of the liver; 18 patients underwent lobectomy, 28 patients had segmentectomy, and 52 patients had subsegentectomy. In the 21 non-cirrhotic patients, 11 patients underwent Obectomy, 5 patients had segmentectomy, and 5 patients had subsegmentectomy. The operative mortality rate of patients with cirrhosis was 11% and of patients without cirrhosis was 5%. There was no significant difference in hepatic function tests between survivors and nonsurvivors. Lobectomy of 5 cm, and presence of gross vascular invasion were poor iPhr°gnostic signs in terms of survival rates as well as recurrence rates. Of e 51 patients with tumor recurrence limited to the residual liver, 13 Patients underwent repeat resection, and 23 patients were treated by transcatheter arterial chemoembolization. The survival rates of the Patients undergoing repeat resection were significantly better than those of other groups.
~
With further refinements of operative technique and periOperative management, major hepatic resections, including right and left hepatic trisegmentectomy, can now be performed quite safely [1--4]. Unfortunately, however, this is not true When major hepatic resections are applied to the cirrhotic liver. Since more than 80% of patients with hepatocellular carcinoma (blCC) in Japan have co-existing liver cirrhosis [5-7], limited hepatic resections (segmentectomy and subsegmentectomy) are roost frequently undertaken for HCC in our country [8-11]. Our experience with 118 hepatic resections for HCC is summarized in this report. We emphasize the safety and effectiveness of limited hepatic resection of the cirrhotic liver in the treatment of HCC. We also stress the merit of pre-operative aSSessment of hepatic functional reserve. N Reprint requests: Masahiro Suenaga, M.D,, Surgical Service, j~agoya Memorial Hospital 4-305 Hirabari, Tenpaku-ku, Nagoya 468, ~Pan.
Materials and Methods
During a 9-year period from January, 1979 to December, 1987, 118 patients underwent hepatic resection of varying extent for HCC at the Nagoya University Hospital and its affiliated hospitals. One hundred patients were male and 18 patients were female. The age ranged from 25 to 78 years (mean 57.5 - 10.1). Ninety-eight (83.1%) of the 118 patients had coexisting cirrhosis of the liver but with liver function well preserved (compensated cirrhosis). Twenty-six (22.0%) patients were chronic carriers of hepatitis B surface antigen (HBsAg). Pre-operatively hepatic function was assessed by serum albumin, bilirubin, normotest (Hepaplastin test, HPT), plasma clearance rate of indocyanin green (ICG-K), and maximum removal rate of ICG (ICG Rmax) [12]. Parenchymal hepatic resection rate (PHRR) [8] was calculated by computed tomography scan volumetry and residual liver ICG Rmax. Hepatic functional reserve after proposed resection was estimated preoperatively by ICG Rmax of the residual liver. Major hepatic resections (lobectomy or greater) were performed following the techniques described by Starzl and coworkers [1], Subsegmentectomy, which is the resection of a segment of Couinaud [13], was carried out by a method similar to that of Makuuchi and associates [14], and segmentectomy, which is the resection of Healey's segment [ 15], was carried out by a method similar to Takasaki's [16]. The common principle of our surgical techniques was inflow control before resection. Intra-operative ultrasonography was used to identify the lesions and their segmental or subsemental portal triad structures and to secure adequate tumor-free margins (->1 cm), A microwave tissue coagulator [17] was applied to the superficial parenchymal transection to minimize blood loss and ultrasonic surgical apparatus (CUSA) was used for the parenchymal transection since 1984. Eighteen (18.4%) of the 98 patients with cirrhosis of the liver underwent lobectomy (14 right and 4 left), 28 (28.6%) patients underwent segmentectomy, and 52 (53.1%) patients underwent subsegmentectomy or less. Of the 20 patients without cirrhosis of the liver, 11 patients Underwent lobectomy (6 right, 5 left), 5 patients underwent segmentectomy, and 4 patients underwent subsegmentectomy. All of the 118 patients were followed regularly by alpha-
98
Table 1. Operative mortality hepatocellular carcinoma. Operative procedure Patients with cirrhosis Lobectomy Segmentectomy Subsegmentectomy Total Patients without cirrhosis Lobectomy Segmentectomy Subsegmentectomy Total
World J. Surg. Vol. 16, No. 1, Jan./Feb. 1992
following
hepatic
resection
for
No. of patients
Operative deaths (%)
18 28 52 98
6 (33) 2 (7) 3 (6) I l (11)
I1 5 4 20
I 0 0 I
(9) (0) (0) (5)
fetoprotein (AFP) once a month and by ultrasonography and/or CT scan once in every 3 months for at least 2 years and a maximum of 10 years (median follow-up: 971 -+ 617 days). The cumulative survival and recurrence rates were calculated by the method of Kaplan-Meier and the statistical comparisons were made by the method of Cox-Mantel and the Wilcox test. The difference was considered significant when the p < 0.05. Results
Operative Death Any death within 2 months after hepatic resection was considered an operative death in this report. The operative mortality of patients with cirrhosis (11 of 98, 11.2%) and that of HBsAg carriers (4 of 26, 15.4%) was higher than that of patients without cirrhosis (I of 20, 5%) and that of non-HBsAg carriers (8 of 92, 8.7%), respectively, but the differences were not statistically significant. Among the patients with cirrhosis, the operative mortality after lobectomy (6 of 18, 33.3%) was significantly higher than after segmentectomy (2 of 28, 7.1%), and after subsegmentectomy (13 of 52, 5.8%) (p < 0.05 and p < 0.01, respectively) (Table 1). The circumstances of the 12 operative deaths are summarized in Table 2. There were no intra-operative deaths. The earliest death was due to mesenteric vein thrombosis in a patient with cirrhosis on postoperative day 6. Five patients died from postoperative hemorrhage and subsequent hepatic failure from 8 days to 25 days after operation. Five other patients, including 1 patient without cirrhosis, died from hepatic failure from 9 days to 40 days postoperatively. One patient died from massive varicera] bleeding on postoperative day 30. The results of pre-operative hepatic function tests, parenchymal hepatic resection rate (PHRR), and intra-operative blood loss were compared in the operative survivors (2 month survivors) and non-survivors. The data are shown in Table 3 for patients with cirrhosis and in Table 4 for patients without cirrhosis. The data are also shown in Figure 1. None of the hepatic function tests could be related to the operative mortality or 2 month survival rate after hepatic resection. However, the PHRR of operative survivors after Iobectomy was significantly lower (p < 0.01) than that of non-survivors (40.8 --- 13.4% vs 63.2 -+ 11.2%). Also the operative blood loss of survivors after lobectomy and subsegmentectomy was significantly less than that of the non-survivors (p < 0.01).
Survival The survival rates after hepatic resection (including operative deaths) were compared between the 98 patients with cirrhosis and the 20 patients without cirrhosis. The 1 year, 3 year, and 5 year survival rates of the patients with cirrhosis were 74.5%, 55.8%, and 33.0%, respectively, and those of the patients without cirrhosis were 75.0%, 39.4% and 39.4%, respectively. The differences were not statistically significant. Survival rates of patients with cirrhosis are compared among the various extents of hepatic resection in Figure 2. The 1 year, 3 year, and 5 year survival rates were 57.9%, 36.8% and 20.0% following lobectomy, 82.8%, 82.8% and 57.6% following segmentectomy, and 72.0%, 46.2% and 24.0% following subsegmentectomy. The survival rates after segmentectomy were significantly greater than those after lobectomy and subsegmentectomy (p < 0.05). Survival rates of patients with cirrhosis are also compared by the number of gross lesions (single v.s. multiple) in Figure 3, by the largest diameter of the tumor (5 cm) in Figure 4, and by the presence or absence of gross vascular invasion of HCC in Figure 5. Significantly better survival rates were found in patients with single gross tumors, the greatest diameter of 5 cm
~J-~
, 2
i
~
8 .3rs
Fig. 7, Cumulative recurrence rates of a single lesion were significantly lower than those of multiple lesions after hepatic resection (p < 0.05).
Fig. 8. Cumulative recurrence rates o f tumor
World Journal of Surgery O 1992 by the Socit~t6 lnternationale de Chlrur~e
Original Scientific Reports Hepatic Resection for Hepatocellular Carcinoma Masahiro Suenaga, M.D., Akimasa Nakao, M.D., Akio Harada, M.D., Toshiaki N o n a m i , M.D., Yoshikatsu Okada, M.D., H a y a t o Sugiura, M.D., Shinichi Uehara, M.D., and Hiroshi Takagi, M.D. Surgical Service, Nagoya Memorial Hospital, and Department of Surgery II, Nagoya University, School of Medicine, Nagoya, Japan One hundred and eighteen patients underwent hepatic resection for hepatocellular carcinoma from 1979 to 1987. Ninety-eight of these patients had co-existing cirrhosis of the liver; 18 patients underwent lobectomy, 28 patients had segmentectomy, and 52 patients had subsegentectomy. In the 21 non-cirrhotic patients, 11 patients underwent Obectomy, 5 patients had segmentectomy, and 5 patients had subsegmentectomy. The operative mortality rate of patients with cirrhosis was 11% and of patients without cirrhosis was 5%. There was no significant difference in hepatic function tests between survivors and nonsurvivors. Lobectomy of 5 cm, and presence of gross vascular invasion were poor iPhr°gnostic signs in terms of survival rates as well as recurrence rates. Of e 51 patients with tumor recurrence limited to the residual liver, 13 Patients underwent repeat resection, and 23 patients were treated by transcatheter arterial chemoembolization. The survival rates of the Patients undergoing repeat resection were significantly better than those of other groups.
~
With further refinements of operative technique and periOperative management, major hepatic resections, including right and left hepatic trisegmentectomy, can now be performed quite safely [1--4]. Unfortunately, however, this is not true When major hepatic resections are applied to the cirrhotic liver. Since more than 80% of patients with hepatocellular carcinoma (blCC) in Japan have co-existing liver cirrhosis [5-7], limited hepatic resections (segmentectomy and subsegmentectomy) are roost frequently undertaken for HCC in our country [8-11]. Our experience with 118 hepatic resections for HCC is summarized in this report. We emphasize the safety and effectiveness of limited hepatic resection of the cirrhotic liver in the treatment of HCC. We also stress the merit of pre-operative aSSessment of hepatic functional reserve. N Reprint requests: Masahiro Suenaga, M.D,, Surgical Service, j~agoya Memorial Hospital 4-305 Hirabari, Tenpaku-ku, Nagoya 468, ~Pan.
Materials and Methods
During a 9-year period from January, 1979 to December, 1987, 118 patients underwent hepatic resection of varying extent for HCC at the Nagoya University Hospital and its affiliated hospitals. One hundred patients were male and 18 patients were female. The age ranged from 25 to 78 years (mean 57.5 - 10.1). Ninety-eight (83.1%) of the 118 patients had coexisting cirrhosis of the liver but with liver function well preserved (compensated cirrhosis). Twenty-six (22.0%) patients were chronic carriers of hepatitis B surface antigen (HBsAg). Pre-operatively hepatic function was assessed by serum albumin, bilirubin, normotest (Hepaplastin test, HPT), plasma clearance rate of indocyanin green (ICG-K), and maximum removal rate of ICG (ICG Rmax) [12]. Parenchymal hepatic resection rate (PHRR) [8] was calculated by computed tomography scan volumetry and residual liver ICG Rmax. Hepatic functional reserve after proposed resection was estimated preoperatively by ICG Rmax of the residual liver. Major hepatic resections (lobectomy or greater) were performed following the techniques described by Starzl and coworkers [1], Subsegmentectomy, which is the resection of a segment of Couinaud [13], was carried out by a method similar to that of Makuuchi and associates [14], and segmentectomy, which is the resection of Healey's segment [ 15], was carried out by a method similar to Takasaki's [16]. The common principle of our surgical techniques was inflow control before resection. Intra-operative ultrasonography was used to identify the lesions and their segmental or subsemental portal triad structures and to secure adequate tumor-free margins (->1 cm), A microwave tissue coagulator [17] was applied to the superficial parenchymal transection to minimize blood loss and ultrasonic surgical apparatus (CUSA) was used for the parenchymal transection since 1984. Eighteen (18.4%) of the 98 patients with cirrhosis of the liver underwent lobectomy (14 right and 4 left), 28 (28.6%) patients underwent segmentectomy, and 52 (53.1%) patients underwent subsegmentectomy or less. Of the 20 patients without cirrhosis of the liver, 11 patients Underwent lobectomy (6 right, 5 left), 5 patients underwent segmentectomy, and 4 patients underwent subsegmentectomy. All of the 118 patients were followed regularly by alpha-
98
Table 1. Operative mortality hepatocellular carcinoma. Operative procedure Patients with cirrhosis Lobectomy Segmentectomy Subsegmentectomy Total Patients without cirrhosis Lobectomy Segmentectomy Subsegmentectomy Total
World J. Surg. Vol. 16, No. 1, Jan./Feb. 1992
following
hepatic
resection
for
No. of patients
Operative deaths (%)
18 28 52 98
6 (33) 2 (7) 3 (6) I l (11)
I1 5 4 20
I 0 0 I
(9) (0) (0) (5)
fetoprotein (AFP) once a month and by ultrasonography and/or CT scan once in every 3 months for at least 2 years and a maximum of 10 years (median follow-up: 971 -+ 617 days). The cumulative survival and recurrence rates were calculated by the method of Kaplan-Meier and the statistical comparisons were made by the method of Cox-Mantel and the Wilcox test. The difference was considered significant when the p < 0.05. Results
Operative Death Any death within 2 months after hepatic resection was considered an operative death in this report. The operative mortality of patients with cirrhosis (11 of 98, 11.2%) and that of HBsAg carriers (4 of 26, 15.4%) was higher than that of patients without cirrhosis (I of 20, 5%) and that of non-HBsAg carriers (8 of 92, 8.7%), respectively, but the differences were not statistically significant. Among the patients with cirrhosis, the operative mortality after lobectomy (6 of 18, 33.3%) was significantly higher than after segmentectomy (2 of 28, 7.1%), and after subsegmentectomy (13 of 52, 5.8%) (p < 0.05 and p < 0.01, respectively) (Table 1). The circumstances of the 12 operative deaths are summarized in Table 2. There were no intra-operative deaths. The earliest death was due to mesenteric vein thrombosis in a patient with cirrhosis on postoperative day 6. Five patients died from postoperative hemorrhage and subsequent hepatic failure from 8 days to 25 days after operation. Five other patients, including 1 patient without cirrhosis, died from hepatic failure from 9 days to 40 days postoperatively. One patient died from massive varicera] bleeding on postoperative day 30. The results of pre-operative hepatic function tests, parenchymal hepatic resection rate (PHRR), and intra-operative blood loss were compared in the operative survivors (2 month survivors) and non-survivors. The data are shown in Table 3 for patients with cirrhosis and in Table 4 for patients without cirrhosis. The data are also shown in Figure 1. None of the hepatic function tests could be related to the operative mortality or 2 month survival rate after hepatic resection. However, the PHRR of operative survivors after Iobectomy was significantly lower (p < 0.01) than that of non-survivors (40.8 --- 13.4% vs 63.2 -+ 11.2%). Also the operative blood loss of survivors after lobectomy and subsegmentectomy was significantly less than that of the non-survivors (p < 0.01).
Survival The survival rates after hepatic resection (including operative deaths) were compared between the 98 patients with cirrhosis and the 20 patients without cirrhosis. The 1 year, 3 year, and 5 year survival rates of the patients with cirrhosis were 74.5%, 55.8%, and 33.0%, respectively, and those of the patients without cirrhosis were 75.0%, 39.4% and 39.4%, respectively. The differences were not statistically significant. Survival rates of patients with cirrhosis are compared among the various extents of hepatic resection in Figure 2. The 1 year, 3 year, and 5 year survival rates were 57.9%, 36.8% and 20.0% following lobectomy, 82.8%, 82.8% and 57.6% following segmentectomy, and 72.0%, 46.2% and 24.0% following subsegmentectomy. The survival rates after segmentectomy were significantly greater than those after lobectomy and subsegmentectomy (p < 0.05). Survival rates of patients with cirrhosis are also compared by the number of gross lesions (single v.s. multiple) in Figure 3, by the largest diameter of the tumor (5 cm) in Figure 4, and by the presence or absence of gross vascular invasion of HCC in Figure 5. Significantly better survival rates were found in patients with single gross tumors, the greatest diameter of 5 cm
~J-~
, 2
i
~
8 .3rs
Fig. 7, Cumulative recurrence rates of a single lesion were significantly lower than those of multiple lesions after hepatic resection (p < 0.05).
Fig. 8. Cumulative recurrence rates o f tumor
