[Hepatectomy for hepatocarcinoma on a cirrhotic liver: decision plans and principles of perioperative resuscitation. Experience with 204 cases]

Among more than 450 hepatectomies performed in the National Cancer Center Hospital of Tokyo from the beginning of 1977 to the end of 1986, 204 were performed for excision of an hepatocarcinoma on cirrhotic liver. The post-operative mortality (determined until the exit of the patient out of the hospital) was 7.8%. It was rather high at the beginning (20% of the first 30 cases) and progressively decreased with experience, to be only 2.8% of the last 70 cases. According to this experience the authors recommend a decisional diagram which appreciate the quantity of functional liver parenchyma which must be resected (depending on the tumor's characteristics) and the quantity of functional liver parenchyma which can be resected without major operative risk (depending on the gravity of the cirrhosis). They expose their attitude with oesophageal varix, the operative techniques and the peri-operative cares they have trained to decrease dramatically the post-operative complications.

Hepatocellular carcinoma: treatment with intraarterial iodized oil with and without chemotherapeutic agents

Thirty-one patients with hepatocellular carcinoma (HCC) were given either an intraarterial injection of iodized poppyseed oil (Lipiodol) alone (group A, n = 6), an emulsion of iodized oil and doxorubicin hydrochloride (Adriamycin) (group B, n = 15), or chemoembolization with the same emulsion followed by gelatin sponge (Gelfoam) particles (group C, n = 10). Hepatic resection was subsequently performed. The frequencies of complete necrosis of tumor in the main lesion, daughter tumors, tumor thrombus, and foci of intracapsular invasion were evaluated in the cut surface of the resected specimen. Group C demonstrated the best therapeutic effects, showing complete necrosis of the main lesion in 83% (P less than .01), daughter tumors in 53% (P less than .01), tumor thrombus in 17%, and foci of intracapsular invasion in 67%. These results are superior to those reported previously for chemoembolization without iodized oil. Group B showed better results than group A, but the difference was not significant. Iodized oil alone (group A) had practically no therapeutic effect but was helpful in differentiating small HCC from regenerative nodules.

Imaging characteristics of large lipoma and angiomyolipoma of the liver. Case reports

A large lipoma and a large angiomyolipoma of the liver, one case each, are described. In the first case, a 24 X 14-cm homogeneous mass demonstrated a fatty density (-96H) on computed tomography (CT), echogenicity on sonography, and fine neovascular vessels on angiography in the right lobe of the liver. Histopathologic diagnosis of the resected specimen was lipoma. The second case showed a 9-cm fatty mass (-66H) with some enhanced interior nodules in the left lateral area of the liver on CT. Sonography disclosed a large echogenic mass with echolucent nodules, which were seen as stains on the angiogram. Angiomyolipoma was the histopathologic diagnosis. Among these diagnostic modalities, the CT scan provided the most specific and reliable findings. Both patients are doing well more than 2 years and 9 months after resection.

Safety of hemihepatic vascular occlusion during resection of the liver

To minimize both intraoperative bleeding and circulatory and biochemical disturbances due to the interruption of blood flow to the liver, we developed a technique for selective, mostly unilobar, control of the hilar vessels to the liver called hemihepatic vascular occlusion. One hundred and fifty-three patients who underwent hepatic resection were divided into four groups with and without cirrhosis of the liver and with and without hemihepatic vascular occlusion which was limited to 30 minutes followed by five minutes of perfusion, which was repeated if necessary. This technique reduced the intraoperative blood loss and the postoperative hyperbilirubinemia significantly but did not produce larger postoperative change of transaminases or lactic dehydrogenase serum levels when compared with similar resections without vascular control.

Four new hepatectomy procedures for resection of the right hepatic vein and preservation of the inferior right hepatic vein

Among the accessory hepatic veins, the thickest one is the IRHV and is a significant vessel in 20 to 24 per cent of the patients. In these patients, if the right hepatic vein is totally resected, the right posteroinferior area can be preserved with the IRHV. Four types of hepatectomies n which the IRHV was preserved were proposed and three of the four procedures were performed upon five patients. In these five operations, the hepatectomies were successfully performed and blood losses were from 1,020 to 3,200 milliliters. These operative procedures, which have not been described before, could widen the indication of hepatectomy in patients with reduced liver function and tumor involving the right hepatic vein. In order to perform these operative procedures successfully, intraoperative sonography is indispensable.

Necrosis of hepatocellular carcinoma as a result of subintimal injury incurred by hepatic angiography: report of two cases

This report described two patients with hepatocellular carcinoma in whom angiographic procedure caused an inadvertent subintimal injury of the hepatic artery, resulting in tumor necrosis. In the first case of a 38-year-old male, complete obstruction of the common hepatic artery occurred on the initial angiography. It was followed by marked reduction of the tumor vessels on repeat angiography, and necrosis of about half of the tumor as confirmed by computed tomography. In the other 58-year-old female, severe subintimal injury occurred in the proper hepatic artery followed by obstruction of the feeding arteries. Subsequent computed tomography scan disclosed necrosis of the tumor. Both patients presented the postembolization syndrome that consisted of a transient fever and elevation of blood enzymes. When spontaneous regression or reduction of hepatocellular carcinoma is observed, special attention should be paid regarding whether or not hepatic angiography was performed and clinical symptoms followed it.

Clinical and radiologic features of hepatocellular carcinoma originating in the caudate lobe

Five patients with hepatocellular carcinoma in the caudate lobe were evaluated. Computed tomography (CT) scan and/or angiography clearly demonstrated multiple intrahepatic metastases in four (80%), and tumor thrombi in the portal vein in two (40%), and in the inferior vena cava in one. Even though there was no recognizable lung metastasis, metastases were found in the orbita in one patient, and in the ribs and thoracic vertebrae in two patients. Four patients died after a mean period of 5.5 months from the initial diagnosis. The mechanism for early invasion into the vessels and multiple intrahepatic metastases of hepatocellular carcinoma arising from the caudate lobe is discussed.

Angiography of small hepatocellular carcinomas: analysis of 105 resected tumors

One hundred five hepatocellular carcinomas less than 5 cm in diameter were resected in 75 patients. The tumors were studied with respect to their detection rate by angiography and their angiographic features. Angiography identified 86 (82%) of the 105 lesions, missing 19 (18%). The findings included tumor vessels (70%) and tumor staining (76%). Pathologic analysis of the 19 undetected lesions showed that 74% of them were smaller than 2 cm in diameter and that they were well-differentiated carcinomas. Forty percent of 100 lesions were in the anterosuperior subsegment of the right lobe.

[Surgical treatment of hepatocellular carcinoma]

The result of surgical treatment for hepatocellular carcinoma in National Cancer Center Hospital is reported. Almost 80% of the cases in this series are cirrhotic. The cases were divided into two groups according to the period when hepatectomy was done, as the first half, [A], was from 1974 till 1981 when the treatment was only surgical removal and the latter half, [B], was from 1982 till the end of March, 1986 when pre- and post-operative embolization was combined with hepatectomy. The operative mortality rate was 10.1% (9/89) for [A] and 2.2% (3/121) for [B]. The cumulative survival rates of [A] for the 1st-5th year were 65.8%, 52.6%, 39.5%, 32.9% and 27.2%, respectively, and those of [B] were 86.1%, 78.1%, 65.1%, 50.6% and 50.6%.

Hepatic lobar atrophy following obstruction of the ipsilateral portal vein from hilar cholangiocarcinoma

Gross deformity of the liver associated with hilar carcinoma is rare. In 17 patients with hilar cholangiocarcinoma and intrahepatic bile duct dilatation, the relationships between lobar or segmental atrophy, compensatory hypertrophy, and patency of portal vein branches were evaluated with computed tomography (CT) and angiography. All six patients with obstructed or narrowed portal veins (group A) had lobar or segmental atrophy on CT scans and angiograms. Compensatory hypertrophy was observed in the unaffected lobe with a patent portal vein in five. In contrast, neither hepatic atrophy nor hypertrophy was demonstrated in the other 11 patients with patent portal veins. All group A patients had differences in hepatic attenuation on CT scans or dense opacification during the hepatogram phase of angiography. Biliary decompression was optimized when the bile duct selected for percutaneous drainage paralleled a patent portal vein. Knowledge of radiologic findings will assist in determining the primary site along the bile duct from which carcinoma has arisen.

Atypical radiographic findings in hepatic cavernous hemangioma: correlation with histologic features

Three cases of cavernous hemangioma of the liver having atypical imaging features are described. In the first case, an 11 x 10 cm hypervascular mass with a central hypovascular area supplied by many large tortuous vessels was demonstrated on angiography, CT, and sonography. The mass had a central cavity that contained fluid. In the second case there was a 2-cm isoechoic mass with a peripheral halo, despite the presence of typical features of hemangioma on angiography. The halo was shown histologically to be fibrous tissue surrounding the tumor. In the third case, a 6.5 x 4.3 cm mass exhibited slight peripheral enhancement on CT but was not detectable by angiography. Histologically, the tumor was hyalinized with only scattered small foci of hemangioma within it.

[A case of cholangiocarcinoma and dysgerminoma associated with Turner's syndrome]

A rare case of cholangiocarcinoma and dysgerminoma synchronously associated with Turner's syndrome was reported. A 53-year-old woman was admitted to our hospital on July 12, 1984 due to intrahepatic and left inguinal tumors. Physical examination revealed the typical characteristics of Turner's syndrome. The karyotype from myelocyte and fibrocyte culture was interpreted as 45, X. The resected intrahepatic tumor was cholangiocarcinoma that had invaded to the transverse mesocolon and the duodenum, while the inguinal lesion was dysgerminoma derived from a dysgenetic gonad. As far as we have been able to ascertain through our investigations of the literature on malignant disease in Turner's syndrome, the association of cholangiocarcinoma with Turner's syndrome has not been reported previously.

Preoperative imaging of liver metastases. Comparison of angiography, CT scan, and ultrasonography

Thirty-one patients with mostly colorectal cancer metastases to the liver had preoperative selective/superselective angiograms (24 cases), computed tomography (CT) [26 cases, mostly enhanced by contrast administered by a peripheral vein (9), the common hepatic artery (9), or the portal vein (5)], and ultrasonography (26 cases). Intraoperative ultrasonography and palpation and examination of the resected specimens revealed 113 tumors. CT detected almost half of the masses smaller than 1 cm, and ultrasonography and angiography about one-third of lesions 1-2 cm in size. Ultrasonography was less powerful for examination of the posterior segment of the liver. CT and ultrasonography placed the tumors into subsegments more accurately than did angiography. Almost 40% of the preoperative plans had to be changed: in two-thirds by extended resections and in one-third by a change from curative to palliative intent. Most changes were due to extrahepatic tumor growth, often within areas screened before surgery. The use of all three imaging modalities for liver metastases is recommended for preoperative planning.

Ultrasonically guided subsegmentectomy

A new operative procedure for systematic subsegmentectomy guided by ultrasound has been described. This operation consists of operative sonography, ultrasonically guided puncture and injection of dye and hemihepatic blood occlusion. Systematic subsegmentectomy was performed upon 57 patients without operative mortality. The cumulative one year survival rate of 35 patients with hepatocellular carcinoma who underwent operation at our hospital was 80.3 per cent. The two and three year survival rates were 63.3 and 52.6 per cent, respectively.

Ultrasonically guided subsegmentectomy

A new operative procedure for systematic subsegmentectomy guided by ultrasound has been described. This operation consists of operative sonography, ultrasonically guided puncture and injection of dye and hemihepatic blood occlusion. Systematic subsegmentectomy was performed upon 57 patients without operative mortality. The cumulative one year survival rate of 35 patients with hepatocellular carcinoma who underwent operation at our hospital was 80.3 per cent. The two and three year survival rates were 63.3 and 52.6 per cent, respectively.

Bile duct necrosis: complication of transcatheter hepatic arterial embolization

Bile duct necrosis because of transcatheter hepatic arterial embolization (THAE) in two patients with hepatocellular carcinoma is reported. Preoperative THAE was performed on 29 patients, and bile duct necrosis was experienced by two of the 29 (7%). In these two patients, gelatin (Gelfoam) powder was used as the embolus. Among the 24 whose embolus was clear, four were embolized with gelatin powder. Therefore, incidence of bile duct necrosis after THAE with gelatin powder was 50%. Because of the hazards of severe complications such as bile duct necrosis, we conclude that gelatin powder should not be used except for the THAE of no more than one segment of the liver.

[Clinical studies on changes in serum transaminase, lactate dehydrogenase, total bilirubin and alkaline phosphatase levels after hepatectomy with and without the hemihepatic vascular occlusion technique]

Changes of serum glutamic oxaloacetic transaminase (GOT), glutamic pyruvic transaminase (GPT), lactate dehydrogenase (LDH), total-bilirubin (T-Bil) and alkaline-phosphatase (AL-P) before operation and for one week of the postoperative period were studied in 45 patients (32 cirrhotic, 13 non-cirrhotic patients) who underwent hepatectomy with the hemihepatic vascular occlusion technique, and compared with 108 patients (42 cirrhotic, 66 non-cirrhotic patients) without it. The blood loss during hepatectomy with hemihepatic vascular occlusion markedly decreased 1500 ml on the average, compared with non-occluded patients. Serum GPT, GOT and LDH level elevated in several postoperative days, however, there was no difference between both groups. Serum total bilirubin level of patients without hemihepatic vascular occlusion elevated more than those with it. This difference was statistically significant. This tendency was more evident in the patients with cirrhosis. Serum AL-P level of patients without hemihepatic vascular occlusion decreased statistically less than those with it. This tendency was more prominent in patients with cirrhosis. With these results, there was no evidence of augmenting the postoperative liver damage by hemihepatic vascular occlusion, even in the patient with cirrhosis. Moreover, the elevation of postoperative serum total bilirubin level was suppressed by hemihepatic vascular occlusion because of the minimum blood loss and minimum blood transfusion.

Hepatectomy for metastatic liver tumor

Hepatectomy for liver metastasis from the colo-rectum and other organs was performed in 77 cases at the National Cancer Center Hospital. The operative procedures included every variety of hepatectomy from extended resection as trisegmentectomy to small resection as partial resection. The operative mortality rate was 3.9% (3/77). The overall cumulative 5-year survival rate was 26.7 +/- 13.9% (survival rate +/- Greenwood's 5% standard error). The long-term survival of the patients with unilobular metastasis (H1) was better than that of bilobular (H2) and disseminated (H3 metastasis). When H1 was divided into solitary unilobular metastasis (H1 solitary) and multiple unilobular metastasis (H1 multiple), the prognosis for the former was better than for the latter after hepatectomy. The long-term survival of patients with H1 multiple was the same as that of patients with H2 and H3; that means that one of the factors affecting prognosis after hepatectomy is whether the tumor is solitary or multiple.

Intrahepatic venous collaterals forming via the inferior right hepatic vein in 3 patients with obstruction of the inferior vena cava

When the inferior vena cava is obstructed, collateral veins enlarge, connecting with the inferior (accessory) right hepatic vein (IRHV) and thence through various hepatic veins to the right atrium. Three such cases are described. In one patient, most contrast material flowed into the IRHV and from there to the left hepatic vein. The second patient had several large collaterals arising from the IRHV and flowing into the right and middle hepatic veins, while the third patient demonstrated anastomoses between the IRHV and the middle hepatic vein. All of these hepatic venous shunts eventually drained into the right atrium. There were no clinical manifestations such as ascites, edema, or dilatation of the abdominal veins. Cavography alone or combined with computed tomography proved to be diagnostic in the assessment of these intrahepatic collaterals.

Primary Budd-Chiari syndrome: ultrasonic demonstration

Three cases of primary Budd-Chiari syndrome were found by ultrasonic examination. These were confirmed by hepatic venography and inferior vena cavography. The ultrasound findings in these patients included communicating vessels between hepatic veins, enlarged inferior right hepatic vein, reversed blood flow in the hepatic vein, and obstruction of the inferior vena cava. With these findings, ultrasound can be used to diagnose primary Budd-Chiari syndrome without hepatic venography or inferior cavography.

Ultrasonically guided cholangiography and bile drainage

Ultrasonically guided percutaneous transhepatic cholangiography (UG-PTC), bile drainage (UG-PTBD) and gallbladder drainage procedure (UG-PTGBD), developed by us, were performed in 47, 183 and 36 patients, respectively. In 47 patients UG-PTC was successfully performed 51 times without complications. By UG-PTBD 220 intubations were carried out successfully and four attempts failed (1.8%). The main complication was that the catheter slipped out from the bile duct. It was experienced 27 times (12.3%) in 23 patients (12.4%) from two to 47 days after intubation. UG-PTGBD was successfully performed 36 times. Bleeding from the catheter was experienced in four patients. However, other complications such as cholascos were not experienced. Due to the development of ultrasonic diagnosis and the UG-PTBD procedure, the indications for percutaneous transhepatic cholangiography (PTC) are now limited. For differentiation of jaundice, ultrasonic examination takes over from PTC. For preparation of PTBD, thin needle cholangiography is no longer necessary because UG-PTBD is a single-step procedure without the need for cholangiography. Therefore, the indication for PTC is limited to patients with partial dilatation of intrahepatic bile ducts without jaundice, for example when only the left hepatic duct is dilated due to hepatolithiasis.

[Ultrasonic diagnosis of liver tumors--with special reference to its limitations in predicting histology]

Since 1980, ultrasound has become the main diagnostic modality in detecting small hepatocellular carcinomas. We laparotomized 72 patients with 93 hepatocellular carcinomas less than 5 cm in diameter and resected 86 tumors in 68 patients. Sensitivities for these small hepatocellular carcinomas were: ultrasound 90.1%, angiography 86.8%, and CT 91.3%. Ultrasound was the most excellent diagnostic modality in diagnosing the location of the tumor, because of its capability of demonstrating the portal and the hepatic venous branches simultaneously. The characteristics of small hepatocellular carcinomas were a mosaic pattern of internal echos, posterior echo enhancement, thin halo and lateral shadows. These findings were recognized in 68%, 61%, 37% and 18% of cases respectively. However, most of the tumors less than 2 cm appeared as simple low-echo areas. In tumors measuring from 2.1 to 5 cm, these findings were demonstrated in 77%, 74%, 48% and 23% of cases respectively. Liver tumors with a mosaic pattern and posterior echo enhancement were seen only in case of hepatoma. Predictability of hepatocellular carcinoma with these two findings was 60% in patients with hepatocellular carcinomas 2.1 to 5 cm in diameter. Predicting the histology of liver tumors less than 2 cm in diameter was difficult with any diagnostic modalities. However, since small hemangiomas less than 2 cm in diameter were delineated as characteristic echogenic areas, there were patients in whom only ultrasound could predict the tumor histology, even if other diagnostic methods could not. In order to diagnose the histology of small liver tumors, ultrasonically guided aspiration cytology has been generally accepted as a reliable procedure. However aspiration cytology made correct diagnoses in about half the tumors which could not be predicted with any other diagnostic modalities. Among hepatomas less than 2 cm in diameter, those due to low grade malignancy hepatocellular carcinomas are very common (33%). Needle biopsy of the tumor is therefore indispensable.

Spontaneous portal-hepatic venous shunt via an intrahepatic portal vein aneurysm

In this unusual case, a portal-hepatic venous shunt was first demonstrated as an abnormal lesion in the liver using a contrast-dynamic computed tomographic scan. It was subsequently shown by percutaneous transhepatic portography to be an intrahepatic portal vein aneurysm communicating with a hepatic vein.