Cases in China (including 2 patients in our series and 8 cases reported in the English literature)
Combined with the 2 patients in our series, 76 patients in China have been reported (Table ). Among the 76 patients, 50 were men, with a male to female ratio of 1.9:1. The mean age of the 71 patients which have a clear age or mean age record was 51.0 years old (range, 10-77 years), and 88.1 percent (52/59) of the patients were older than 40 years.
Clinicopathological characteristics of primary hepatic malignant fibrous histiocytoma
Most of hepatic MFH were large neoplasms, and only 13.2 percent (10/76) were smaller than 5 cm. The exact sizes of hepatic MFH were mentioned in 59 cases, including 19 cases with diameter descriptions in radiological investigations. The mean diameter was 10.3 cm (range, 2.5-23.5 cm).
Among the 76 cases, 82.9 percent (63 cases) of the tumors were described as solitary lesions, while only 17.1 percent of the tumors were composed of multiple lesions. The information of tumor localization was provided in 66 cases, including 10 cases with multiple tumor nodules. Among the 10 cases, the tumor nodules of 4 cases were found in both right and left lobes and the tumor nodules of the remaining 6 cases were confined to one lobe of the liver, 3 in the left and 3 in the right lobe. Among the 56 cases with a solitary lesion, tumors were often found in the right lobe (33 cases), and less frequently in the left lobe (21 cases), and the lesions of the remaining 2 cases (3.6%) were in the middle part of the liver.
For clinical manifestation of hepatic MFH, 74 cases provided details. The symptoms were usually non-specific. 58 patients (78.4%) complained of abdominal pain or discomfort of different degrees, accompanied with loss of body weight in 12 cases, obstructive jaundice in 1 case, nausea and anorexia in 8 cases, and fever in 15 cases, including 2 cases with the temperature more than 39 degree. Nearly 14.9 percent (11 cases) of the patients were asymptomatic, with the tumors found during routine physical examinations or unrelated conditions. In addition, 2 patients were admitted to hospital for obstructive jaundice and 2 patients were for fever. The last one patient was admitted for loss of body weight and the lesion was found by the radiological investigation on admission.
A variety of radiological investigations were performed for the evaluation, including ultrasonography (US, 46 cases) and computed tomography (CT, 63 cases) in the vast majority of the patients, and magnetic resonance imaging (MRI) in only 16 patients. The tumors were described on cross-sectional imaging as solid, cystic or heterogeneous (solid and cystic) masses. Among the US investigations of 46 cases, 34 cases provided details. The tumors appeared hypoechoic in 10 cases, hyperechoic in 10 cases, anechoic in 2 cases, and showed a mixed pattern that may include extensive necrotic areas in 8 cases. In addition, the tumors were described as cyst-like lesions in 2 cases. Among the remaining 2 cases, the lesions of one case were not detected and the other case had a lesion in each lobe of the liver, a cyst-like lesion in the left lobe and a hypoechoic lesion in the right. We have detail CT descriptions of 49 cases. The lesions that were not detected in the US investigation were not detected either in the CT investigation. Another case had one lesion in each lobe of the liver, a hypodense mass in the left lobe and a cyst-like lesion in the right. Among the remaining 47 cases, on plain CT scanning, MFH appeared as hypodense masses with necrotic areas of differing degrees in 40 cases, cyst-like masses in 4 cases, mixed masses in 3 cases. 45 cases provided details on the enhancement scanning. The solid components or the cystic walls of 41 cases were introduced to be variably enhanced, from slightly to markedly, while the masses of the remaining 4 cases showed no enhancement. Among the 16 cases with MRI examinations, 2 cases provided no information of T1-weighted MR images. The T1-weighted MR images of the remaining 14 cases all showed low signal-intensity masses. T2-weighted MR images of all the 16 cases showed high signal-intensity solid masses. 11 cases provided details on contrast-enhanced dynamic MR images. A gradual enhancement of internal components was present in 10 cases, and no obvious enhancement was showed in only 1 case.
For laboratory examinations, 5 cases provided no details. It showed that 77.8 percent (35/45) of liver function tests and 60.8 percent (31/51) of hepatitis investigations were normal in the patients. About the tumor marker, only a few of patients have slightly elevated tumor marker levels, such as serum alpha-fetoprotein (AFP, 7/68), carcinoembryonic antigen (CEA, 1/41), carbohydrate antigen 19-9 (CA19-9, 3/34). 5 cases have the description of carbohydrate antigen 125 (CA125) examinations and only 1 patient has an elevated level.
Despite the technological advances, pre-operative diagnosis of hepatic MFH is still difficult, and the rate of misdiagnosis is extremely high. We had information about preoperative diagnosis for 67 patients, none of which, including one case with puncture biopsy before operation in our series, got the exact diagnosis. 48 cases were diagnosed as hepatic neoplasms, 8 cases were only diagnosed as masses of liver. In addition, 10 cases (14.9%) were even misdiagnosed as benign lesions of the liver, including hemangioma in 3 cases, hepatophyma in 5 cases, hepatic hydatid cyst in 1 case and hepatic cyst in 1 case. The remaining 1 case was only diagnosed as calculi of intrahepatic duct, and the hepatic mass was not discovered until surgery.
Surgical procedure was the main treatment for hepatic MFH (Table ). Among the 68 patients whose treatments were available, 49 cases (72.1%) had integrated resection, 4 cases (5.9%) had palliative operation (local resection), 7 cases had (10.3%) puncture biopsy, and the other 8 cases (11.8%) were found unresectable at laparotomy and underwent biopsy. In one case with integrated resection, a lesion in the lung was also detected synchronously with the one in the liver, and left pneumonectomy was performed 50 days after operation. The lesion of the lung was demonstrated as metastatic MFH and no adjunctive therapy was performed afterward. Among the remaining 48 vases with integrated resection, 3 cases received chemotherapy, 2 cases received radiation therapy, 1 case received hepatic artery and portal vein subcutaneous embedding alternate infusion chemotherapy after operation, 1 case received hepatic arterial infusion chemotherapy twice because of tumor recurrence 2 months after operation, and 1 case received lymphadenectomy and 125I radiation therapy because of lymph nodes metastasis 3 months after operation. Among the 4 cases with palliative operation, 1 case received intrahepatic dehydrated alcohol injection and hepatic arterial chemoembolization six times and regular antituberculosis treatment because misdiagnosised as intrahepatic nonspecific inflammatory granuloma, HCC or tuberculosis respectively before operation, and after operation superselective arterial chemoembolization and hepaticarteriography chemoembolization were also performed, another case received drainage fenestrated with laparoscope because misdiagnosised great cyst in the left lobe of liver before laparotomy, and received palliative operation because of intraperitoneal plant. Among the 7 cases with puncture biopsy, 1 case received percutaneous cryoablation combined with transarterial chemoembolization. Among the 8 cases with biopsy after laparotomy, 2 cases received chemotherapy.
Treatment and prognosis of primary hepatic malignant fibrous histiocytoma
Most cases of primary hepatic MFH were the type of pleomorphic storiform MFH. One case in the Chinese literature was the type of inflammatory MFH, which has not been reported in the English literature. In the English literature, the only case that was diagnosed as giant cell MFH was also from China. For immunohistochemical investigations, only 28 cases provided details. Immunohistochemically, the tumor cells showed typically positive for Vimentin(27/27), CD68(13/13), and MAC387(2/2), and AAT(13/14), Lysozyme(11/12), ACT(10/12) showed positive in most cases. AFP(5/5), FVIII(2/2), CEA(4/4), EMA(9/9), CK8/18(7/7), Hep parl(4/4), CD10(3/3), HMB-45(7/7), Actin(5/5), Desmin(13/13) showed all negative, and CK(1/12), CD34(1/5), S-100(1/10), SMA(1/5) were expressed only in a few tumors.
To our knowledge, follow-up periods ranging from one month to 10 years were mentioned for 42 patients. The 8 cases with no treatment introduction have no follow-up either. Among the 49 cases with integrated resection, a postoperative follow-up was conducted in 33 patients. 10 cases of these 33 patients had recurrence or metastasis within a year, 2 cases had recurrence or metastasis over a year, 5 cases had no recurrence or metastasis within a year, 2 cases had no recurrence or metastasis over a year, 8 cases died within a year, and the remaining 6 cases were still alive with no detail information provided (the follow-up periods of these 6 cases was 1 month, 3 months, 23 months, 26 months, 3 years and 10 years respectively). Among the 4 cases with palliative operation, 2 cases were followed-up. 1 case died of multiple organ failure 2 months after operation, and the other was lost after finding the lesion was increasing 4 month after operation. Among the 7 cases with puncture biopsy, 2 cases were followed-up, and died 1 month and half of 1 month respectively after discharge from hospital. Among the 8 cases with biopsy by operation, 5 cases were followed-up. 4 cases died 1 month, 2 month, 3 month and half of 1 year respectively after discharge from hospital, and 1 case died without the surviving time provided.