37 Malignant liver tumours
Page:
1 Historical review 772
2 Classification 772
2.1 Systematization 772
2.2 TNM staging system 773
3 Hepatocellular carcinoma 774
3.1 Definition 774
3.2 Epidemiology and frequency 774
3.3 Risk factors and causes 774
3.4 Pathogenesis 776
3.5 Morphology 776
3.5.1 Macroscopic forms 776
3.5.2 Microscopic forms 777
3.5.3 Cytological differentiation 777
3.5.4 Metastatic spread 777
3.6 Symptomatology 777
3.6.1 Subjective complaints 777
3.6.2 Clinical findings 778
3.7 Diagnosis 778
3.7.1 Laboratory findings 778
3.7.2 Imaging procedures 779
3.7.3 Morphological diagnosis 780
3.8 Prognosis 781
3.9 Therapy 782
3.9.1 Surgical therapy 782
3.9.2 Interventional therapy 783
3.9.3 Medicinal therapy 786
3.9.4 Adjuvant measures 786
3.10 Fibrolamellar carcinoma 788
4 Cholangiocellular carcinoma 788
4.1 Definition 789
4.2 Morphology 789
4.3 Epidemiology 790
4.4 Pathogenic risk factors 790
4.5 Clinical features and diagnosis 790
4.6 Therapy 791
4.6.1 Invasive and surgical techniques 791
Page:
4.6.2 Chemotherapy and radiotherapy 791
4.6.3 Adjuvant measures 791
5 Cystadenocarcinoma 791
6 Hepatablastoma 792
7 Mesenchymal liver tumours 792
7.1 Embryonal sarcoma 792
7.2 Epithelioid haemangioendothelioma 792
7.3 Angiosarcoma 793
7.4 Leiomyosarcoma 793
7.5 Rhabdomyosarcoma 793
7.6 Fibrosarcoma 794
7.7 Fibrous histiocytoma 794
7.8 Malignant schwannoma 794
7.9 Hepatic liposarcoma 794
7.10 Hepatic osteosarcoma 794
8 Neuroendocrine tumours 794
8.1 Hepatic gastrinoma 794
8.2 Primary hepatic carcinoid 795
9 Malignant lymphoma 795
10 Liver metastases 795
10.1 Definition 795
10.2 Morphology 795
10.3 Symptomatology 796
10.4 Diagnosis 797
10.5 Metastases in children 799
10.6 Therapy 799
10.6.1 Resection therapy 800
10.6.2 Cryotherapy 800
10.6.3 Liver transplantation 800
10.6.4 Local treatment 800
10.6.5 Local chemotherapy 801
10.6.6 Systemic therapy 801
10.6.7 Systemic chemotherapy 801
10.6.8 Adjuvant measures 802
앫 References (1⫺364) 802
(Figures 37.1 ⫺37.30; tables 37.1⫺37.12)
1 Historical review
䉴
C. A. R okitansky (1849)
was probably the first author to refer to primary liver carcinoma as an independent disease. Until then, the problem had been to accept the existence of both primary and secondary (metastatic) liver tumours and to differentiate between them histologically.•
In 1854E. N oeggerath
described a congenital hepatic carcinoma as being a mechanical obstetric obstacle. In 1859T. B illroth
reported the presence of hepatic metastases from a cylindroma.A. K elsch
andB. L. K liener
pre- sented the first case reports of a primary liver tumour in 1876.C. S abourin (1881)
made significant histological progress by dif- ferentiating between hepatocellular and cholangiocellular carci- nomas. He reported on four more patients suffering from pri- mary hepatic carcinoma and coined the term “hepatoma”.J. A. P. P rice (1883)
described the development of carcinoma from cirrhosis, which led to the introduction of the term “cirrhosishepatis carcinomatosa”. V. C. H anot et al. (1888)
attempted to dif- ferentiate liver tumours according to macroscopic and micro- scopic criteria.• H. E ggel (1901)
confirmed the association between cirrhosis and liver carcinoma, actually finding cirrhosis in 85% of all liver carcinomas. He distinguished carcinomas according to nodular, massive or diffuse growth, evaluating all cases that had been published since 1865. (5) Primary cystade- nocarcinoma of the liver was differentiated for the first time in 1909 (P. B ascho
). To our knowledge, the first right-sided lobec- tomy for hepatic carcinoma was carried out in 1911 (W. W en- del
).•
In 1932T. Y oshida
succeeded in creating a hepatocellular carcinoma in animal experiments by using o-amidoazotoluol:this marks the beginning of the search for chemotoxicological causes.
M. M. S teiner (1938)
wrote a detailed report on primary liver carcinoma in children. (239)L. L isa et al. (1942)
observed that metastases were rarely found in a cirrhotic liver. In 1952C. B erman
presented an overview of about 2,000 cases of pri- mary liver cell carcinoma which had been reported up to that time.•
In 1950G. M. F indlay
postulated an association with chronic viral hepatitis, which he had detected in his epidemio- logic examination of soldiers stationed in regions south of the Sahara desert, a finding that was also reported byM. P ayet et al.
in 1956. This supposition was confirmed when the hepatitis B virus was discovered(J. B. S mith et al., 1965
;S. S herlock et al., 1970)
. (4, 13, 151)2 Classification
2.1 Systematization
Malignant liver tumours may be grouped as follows: (1.) primary forms, i. e. originating in the liver (s. figs. 29.14;
30.2; 37.1, 37.7, 37.8), (2.) secondary forms, i. e. arising from metastases (s. figs. 37.2, 37.16, 37.17, 37.21, 37.28 ⫺30), and (3.) extrahepatic tumours infiltrating the liver from the outside (e. g. gall-bladder carcinoma).
(s. fig. 37.3)
Primary liver tumours originate in principle from all his- togenetic cell elements found in the liver: hepatocytes, bile-duct epithelia, periductal biliary glands, neuroendo- crine cells and mesodermal structures (such as endothe- lial cells, Ito cells, Kupffer cells) as well as fibroblasts,
Fig. 37.1: Grey-coloured, medium-coarse tubercular cirrhosis in
haemochromatosis with hepatocellular carcinoma: in the fore- ground, white, flat tumour granuloma of the right hepatic lobe with vascularization at the tumour margin and small “cancer umbilicus”;in the background, two additional tumour granulomas. Carcinomas infiltrating the peritoneal serosa in the right upper abdomen
Fig. 37.2: Liver metastases in the right hepatic lobe in breast
cancer with pronounced chaotic vascularization. Dissipated light reflex due to tumourous tubercles on the surfacenerve cells and muscle cells. However, they may also occur as mixed forms. In rare cases, ectopic tissue in the liver may be the starting point for malignant tumours.
(s. tab. 37.1)
Primary liver tumours can be categorized according to
macroscopic criteria. The solitary coarse-granulomatous
type is predominantly found in the right lobe, while the
nodular multi-granulomatous type is most common in
cirrhosis. The diffuse infiltrative type is relatively rare
Fig. 37.3: Gall-bladder carcinoma subsequent to cholelithiasis with
chronic cholecystitis and shrunken gall bladder. Bulbous carci- noma infiltrating the right lobe of liverPrimary liver tumours Epithelial tumours
1. Hepatocellular tumour
⫺ spindle-cell HCC
⫺ fibrolamellar carcinoma 2. Cholangiocellular tumour
⫺ mucoepidermoid carcinoma 3. Combined HCC / CCC 4. Biliary cystadenocarcinoma 5. Hepatoblastoma
Mesenchymal tumours
1. Embryonal sarcoma
2. Malignant epithelioid haemangioendothelioma 3. Angiosarcoma
4. Leiomyosarcoma 5. Rhabdomyosarcoma 6. Fibrosarcoma 7. Fibrous histiocytoma 8. Malignant schwannoma 9. Hepatic liposarcoma 10. Hepatic osteosarcoma
Malignant lymphoma Neuroendocrine tumours
1. Hepatic gastrinoma 2. Hepatic carcinoid
Mixed tumours
1. Mixed liver tumour 2. Carcinosarcoma
Secondary liver tumours
1. Metastases 2. External infiltration
Tab. 37.1: Systematization of primary and secondary malignant
liver tumours(5 ⫺10% of cases). A combined type is observed on occa- sions. The right lobe is affected seven times more fre- quently than the left lobe. Some 3 ⫺5% of malignant liver
tumours are of primary origin. • In contrast, 95 ⫺97% of all malignant liver tumours are assumed to be secondary liver tumours, so-called liver metastases. (s. tab. 37.1)
2.2 TNM staging system
In medical letters/reports, an identified tumour is (cor- rectly) classified according to the so-called TNM stag- ing system. In this process, unclear or even incorrect interpretations are sometimes given either due to incon- sistent use of the terminology or lack of sufficient experience with this kind of classification. Thus it is appropriate to explain this system with interpretations.
The anatomical spread of malignant liver tumours is also described according to the TNM staging system. It is extremely important to assess the histologically verifi- able invasion of vessels in the preparation as well as to clarify the question of whether the resected tissue still contains tumourous foci. (s. tab. 37.2)
䊐 T Primary tumour
Tis Preinvasive carcinoma (carcinoma in situ) T1⫺T4 Size and/or local extension of the primary tumour T1 Solitary node < 2 cm, no vascular invasion T2 Solitary node < 2 cm, with vascular invasion
or
Multiple nodes, limited to one lobe, no node > 2 cm, no vascular invasion
or
Solitary node > 2 cm, no vascular invasion T3 Solitary node > 2 cm, with vascular invasion
or
Multiple nodes, limited to one lobe, each < 2 cm, with vascular invasion
or
Multiple nodes, limited to one liver lobe, one node
> 2 cm, with or without vascular invasion
T4 Multiple nodes in more than one liver lobe or a tumour affecting a larger branch of the portal or hepatic vein
Tx Minimum requirements for determining the locus or the degree of extension of the primary tumour are not fulfilled
䊐 N Regional lymph nodes
N0 No regional lymph node metastases N1 Homolateral lymph nodes affected
N2 Bilateral or contralateral lymph nodes affected, or regional lymph nodes almost grown together
N3 Lymph nodes grown together and/or extended Nx No assessment possible
䊐 M Distant metastases M0 No distant metastases M1 Distant metastases Mx No assessment possible
Tab. 37.2: Pretherapeutic/preoperative TNM staging of malignant
tumours (Tumour/Nodules/Metastases). In postoperative morpho- logic TNM staging, T, N and M are preceded by the letter “p”This TNM staging system comprises the following
stages with a respective survival time of 3 years after
resection of HCC:
stage I
ⴝ T1No Mo (83%) stage IV A T
4N, Mo (10ⴚ25%) stage II
ⴝ T2No Mo (70ⴚ75%) stage IV B T, N, M
1stage III A ⴝ T
3No Mo (45ⴚ50%) stage III B ⴝ T
1N
1Mo (20ⴚ40%)
3 Hepatocellular carcinoma
3.1 Definition
Hepatocellular carcinoma (HCC) originates in the hepatocytes; consequently, HCC resembles liver parenchyma in morphology. It may present as highly differentiated, moderately differentiated or undif- ferentiated (anaplastic). The most common (and basic) structure is the trabecular localization of tumour cells around sinusoidal vessels. • Clinically, HCC is considered to be an extremely malignant, rapidly progressing form. Therapeutic measures still remain limited.
3.2 Epidemiology and frequency
HCC is one of the most common malignant primary liver tumours worldwide. Of all malignant tumours, it ranks fifth in frequency in men and eighth in women ( s ⫽ 6%, t ⫽ 3%). Between 500,000 and one million new cases are reported each year. The annual mortality rate is virtually the same as its annual incidence. Its geo- graphical distribution varies greatly and correlates almost 100 per cent with the regional incidence rates of HBV and HCV infection. There are areas showing low incidence with < 5/100,000 inhabitants/year (e. g. Scan- dinavia, Australia, USA., Central Europe), moderate in- cidence with 5 ⫺20/100,000 inhabitants/year (e.g. East/
South-East Europe, Mediterranean region) and high incidence with 20 ⫺150/100,000 inhabitants/year (e.g.
Asia, Africa).
During the course of the 20
thcentury, the statistical fre- quency of HCC rose constantly ⫺ owing to an increase in autopsies and improvements in diagnostic pro- cedures. • In 1910, for example, autopsy records showed HCC in only 0.04% of cases; in 1930 the figure was 0.13%, in 1960 0.23%, and in 1970 0.6 ⫺0.8%. In 1986 the frequency of HCC was given at about 1% of cases based on autopsy findings. The ratio of HCC to liver metastases is about 1 : 50.
HCC affects persons of all age groups. In Asian and African countries, the morbidity peak is reached in ado- lescence or between the ages of 20 and 40 years, corres- ponding to predominantly perinatal or postnatal infec- tion with hepatitis viruses. In countries with a low incidence, the morbidity peak is between the ages of 50 ⫺60 years. • HCC has also been observed in babies and infants.
The HCC gender ratio between men and women is about 3 : 1, but up to 8 : 1 in countries with a high inci- dence. In a cirrhosis-free liver, however, men and women are affected by HCC at the same rate, i. e. the gender ratio of cirrhosis seems to determine the ratio of HCC as well. Androgens are thought to be of aetiopathogene- tic significance, since carcinoma cells have been shown to carry androgen receptors which display qualities favouring growth, and a carcino-protective effect was observed in animal experiments when androgen was withdrawn. Sex hormones appear to be so-called cocar- cinogens.
(24, 44, 48, 83)3.3 Risk factors and causes
Animal experiments: A considerable number of chemical agents
have proved to be directly carcinogenic in animal experiments. This is most likely to apply to humans as well (but as yet no supporting evidence has been presented). Both a linear dosage-effect relation- ship and a respective time-exposure relationship can be observed:the higher the dosage and the longer the exposure, the greater is the likelihood of carcinoma formation. Carcinomatous degenera- tion occurs predominantly in epithelial or mesenchymal tissues. (s.
tab. 37.3)
Azo-compounds
1. 2,2-azonaphthalin 2. 2,2-diamino-1,1-naphthyl3. m-methyl-p-dimethylaminoazobenzene 4. o-aminoazotoluene
5. p-aminoazobenzene
6. p-dimethylaminoazobenzene (butter yellow) 7. p-monomethylaminobenzene
Hydrocarbon compounds
1. 2-aminofluorene 7. Acetylaminofluorene 2. 4-aminostilben 8. Carbon tetrachloride
3. 2-anthramine 9. Ethylurethane
4. 1,2-benzanthracene 10. Methylcholanthrene 5. 1,2,5,6-dibenzanthracene 11. Thioacetamide 6. 3,4,5,6-dibenzkarbazole
Tab. 37.3: Chemical substances displaying carcinogenicity, as has
been demonstrated in experiments or is very probableThe figures and observations relating to the epidemi- ology and frequency of HCC are almost exclusively the result of individual or combined risk factors. The extent of risk regarding HCC correlates with (1.) aetiology, (2.) duration, and (3.) inflammatory activity of the liver dis- ease. In 10 ⫺15% of patients, no risk factor could actu- ally be determined in the development of HCC. (s. tab.
37.4)
Genetic mechanisms: A genetic disposition is suspected
in HCC, as in other malignant tumours. Some 22% of
patients suffering from HCC had other organ tumours
as well.
(135)• Several hereditary metabolic diseases, with
or without cirrhosis, may increase the risk of HCC con-
siderably. These include tyrosinaemia (type I), glycogen-
osis (type I), alpha
1-antitrypsin deficiency, galactos-
aemia, porphyria cutanea tarda, acute intermittent
porphyria and idiopathic haemochromatosis. Patients
1. Hepatitis viruses
HBV, HCV2. Liver diseases
Chronic hepatitis Cirrhosis NASH
3. Mycotoxins or phytotoxins
Aflatoxin (49, 110, 174) Microcystin
Cycasin Ochratoxin
Luteoskyrin Safrol
Maltrozym
4. Nutrition, social drugs
Alcohol (159, 163) Ethionine surplus Betel quid chewing (168) Tobacco smoke B6and choline deficiency
5. Metabolic diseases
alpha1-antitrypsin deficiency (50) Colon polyposis
Galactosaemia
Glycogenosis (type I) (39, 103) Haemochromatosis (27) Neurofibromatosis Porphyria (54)
Tyrosinaemia (type I) (171)
6. Chemical agents
Alkylating substances Nitrose compounds Aromatic amines Vinyl chloride (148)
Azo-compounds
etc.
7. Inorganic substances
Arsenic, asbestos, Cadmium, chromium, Lead, manganese, nickel8. Medication
Androgens, anabolics Methotrexate Contraceptives (53, 169) Methyldopa Cyproterone acetate
9. Ionizing radiation
Thorium (55, 73) X-raysTab. 37.4: Risk factors (carcinogens and cocarcinogens) regarding
hepatocellular carcinoma, which have already been proved (in- cluding some references) (s. tabs. 29.10; 37.3)with diabetes mellitus and obesity
(29, 65)also have a higher risk of developing hepatocellular carcinoma. • HCC has only rarely been associated with Wilson’s dis- ease; copper seems to have a carcino-protective effect.
(172)
(s. tab. 37.4)
Chemicals: A number of chemicals can cause malignant hepatic tumours, including HCC, both in animal experi- ments and in the everyday life of humans. These include vinyl chloride, aromatic amines, nitrogen compounds, polycyclic aromatic hydrocarbons and alkylating sub- stances. Similarly, mycotoxins and phytotoxins (afla- toxin, cycasin, ochratoxin, safrol, etc.), betel quid chew- ing as well as inorganic substances (arsenic, asbestos, cadmium, chromium, etc.) can have a carcinogenic effect. In addition, substances emitting ionizing radia- tion must be mentioned in this context. (s. tab. 37.4) Medication: The aetiological relationship between the intake of certain types of medication and the develop-
ment of HCC is well known. These include methotrex- ate, methyldopa, androgens/anabolic steroids, oestrogen derivatives/oral contraceptives and cyproterone acetate.
(s. tabs. 29.9; 37.4)
Alcohol and tobacco: Alcohol is considered to be a cocarcinogen, i. e. it requires additional noxae (e. g.
tobacco smoke, aflatoxin) to be able to induce HCC as a carcinogen. The combined effects of alcohol and afla- toxin result in a 35-fold increase in carcinogenicity. Afla- toxin may cause a specific type of mutation within the tumour suppressor gene p53 and thus trigger initiation (i. e. the first step in the development of a tumour). The risk of HCC developing with underlying alcoholic cir- rhosis is about 15% or 22.6% (i. e. there is approximately a 4-fold increase in the HCC risk); abstention from alco- hol does not lower the risk of HCC significantly.
(159, 163)Hepatitis viruses: In chronic infections, HBV DNA may become integrated into the genetic material of the hepa- tocytes; this has a strong, time-dependent carcinogenic effect. Alcohol speeds up the incorporation of viral DNA into the host genome. HBV carriers (those suffer- ing from liver disease as well as healthy persons) have an HCC risk which is 200 times higher than that of other people. Integration of viral DNA into the tumour cells is well documented. Integrated HBV DNA shows a number of modifications. Genetic damaging of hepa- tocytes by the integrated viral DNA (also in serolog- ically HBV-negative patients, who may be molecular- biologically positive) is considered to be the initiation of carcinogenesis. In this process, the HBV subtypes apparently behave like the HBV type itself. HBxAg is undoubtedly of pathogenic significance; possibly it also damages p53. HBV protein (and the core protein of HCV) are able to activate NF-kappa B, which is consid- ered to be important in HCC pathogenesis. Vaccination against HBV infection is thought to be an effective pro- phylactic measure against the development of HCC. • Superinfection with HDV has a potentiating effect, resulting in the earlier occurrence of both cirrhosis and HCC ⫺ but not in a higher incidence of HCC. Animal experiments show that HBV can induce HCC without further cofactors. This evidence of the strong carcinoge- nicity of HBV also explains why, for example, HBsAg- positive men have a risk of developing HCC which is 98 times higher than that of an HBsAg-negative control group of the same age. A higher familial frequency has also been observed.
(67, 110, 151, 167)• HCV is equally carcinogenic; in certain regions, its carcinogenicity is actually 2.7 times higher than that of HBV, whereby different molecular mechanisms are responsible for the development of HCC. HCV is not integrated into the hepatocellular genome. Prevalence of HCV antibodies in HCC patients has been shown to be considerable (albeit varying between individual countries), e. g. 27%
vs 4%, 55% vs 10%, 30% vs 1%, and 71% vs 5%. This
corresponds to a 27-fold increase in the risk of HCC.
Genotype 1 b is associated with a higher risk of HCC as well as a decrease in IGF. Simultaneous infection with GBV-C does not increase the HCC risk. The likeli- hood of HCC occurring subsequent to an HCV infec- tion is 21.5% after 5 years, 53.2% after 10 years, and 75.2% after 15 years. HCV cirrhosis is associated with HCC at a malignant transformation rate of 2 ⫺8% per year.
(67, 127, 157, 164, 167)Liver diseases: An association with AIH and with secondary biliary or CDNC-induced biliary cirrhosis has been reported
(94, 116, 155), albeit as a rare event.
Cirrhosis: In 70⫺80% of cases, HCC develops in under- lying (mainly multicentred, coarse-granulomatous) liver cirrhosis, i. e. some 20% of cirrhotic patients suffer from HCC. • Cirrhosis of any aetiology must thus be con- sidered a precarcinogenic factor. • In Europe and the USA, cirrhosis is mostly alcohol-induced. In HCC patients in Africa or Asia, cirrhosis is less common: gen- erally perinatal, postnatal or juvenile HBV infections, most likely involving large numbers of viruses, result in the formation of HCC before cirrhosis can develop.
Chronic viral hepatitis, just like cirrhosis, causes an increase in hepatocellular proliferation. Unless the genetic damage induced by viral DNA can be repaired by enzymes produced in the hepatocytes, it is passed on via mitoses to the daughter cells. An increased DNA synthesis rate within the cirrhosis is accompanied by an elevated risk of malignant transformation. Cirrhosis increases the susceptibility of the hepatic tissue to other carcinogens. A rise in the mitosis rate thus also results in an increase in genetic alterations. This causes onco- gens to be activated, while tumour suppressor genes or mutation repair genes become inactivated.
(19, 44)3.4 Pathogenesis
Development of HCC begins in the small diploid hepatocytes,
which have a higher growth rate. Pathogenesis is a multifactorial event. The sequential course has three main phases: (1.)initiation:
various noxae cause a genetic defect, which can be repaired by endogenous mechanisms and is thus reversible; (2.)
promotion: if
the genetic damage is beyond repair, the initiated hepatocytes are stimulated and mitosis begins, with the result that the genetic dam- age is transferred to the daughter cells; (3.)progression: clonal
expansion of the altered (“malignatized”) cells occurs.Tumour-suppressor gene: The tumour-suppressor gene p53 is
located on chromosome 17. Noxae may cause mutations of this gene, which results in the loss of its suppressor effect. (6, 126)•
An autosomal dominant mutation of p53 is also encountered in humans, so that a tumour can develop during adolescence, the so-
calledLi-Fraumeni syndrome.
Carcinogens are cancer-causing substances which are able to initi-
ate genetic damage of the hepatocytes (or of bile-duct epithelia or sinusoidal endothelial cells) without the assistance of additional noxae.Cocarcinogens require other cocarcinogens or carcinogens
in order to cause malignant transformation.•
Theinteraction
between hepatitis viruses, alcohol, chemical agents, hormones, etc.is a crucial factor in the development of manifest HCC. In this context, constituents of tobacco smoke, nutritional factors (e. g.
choline deficiency, vitamin B6deficiency, aethionine surplus) or
occasionally enhanced fatty acid synthesis (130) may act as cocar- cinogenic factors.
Growth factors (HGF, IGF, etc.) are also of crucial importance in
pathogenesis, as is a compromised immune defence system.Increased expression of angiopoietin 1 and 2 plays an important role in the vascular development of HCC. (165)
Aetiopathogenesis is multifactorial as an interaction of genetic, exogenous and/or endogenous factors. In a molecular-biological context, HCC is regarded as an extremely heterogeneous tumour.
3.5 Morphology 3.5.1 Macroscopic forms
The macroscopic or histologic differences in the morph- ology of individual hepatocellular carcinoma types are of little epidemiological or clinical significance. • Macroscopically, HCC has a whitish-yellow colour, is often permeated by bile and shows a soft consistency.
There are often haemorrhages in the nodes as well as central necrosis. HCC is mainly supplied with arterial blood. Its doubling time is 30 to 400 days, with an average of 120 days. Occasionally, infiltration into the portal vein system or the hepatic veins occurs (s. fig.
6.16), with intrahepatic metastases and thromboses in the branches of the portal or hepatic veins.
(138)HCC may also become manifest simultaneously with cholan- giocarcinoma.
(95)• Three forms (nearly identical with the three forms described by
H. Eggelin 1901)
(5)can be differentiated: (1.) expansive type, (2.) infiltrative type, and (3.) combination type. • A rare form is the pedunculated HCC (only about 100 cases have been reported). It occurs mostly on the underside of the right lobe and protrudes beyond the upper edge of the liver.
This form (occasionally > 1 kg in weight) can be more easily resected, leading to a better prognosis.
(124, 178)(s. tab. 37.5)
Macroscopic forms Microscopic forms
1. Expansive type (ca. 18%) 1. Trabecular type
• solitary coarse-nodular 2. Pseudoglandular type
• multiple coarse-nodular 3. Scirrhous type 4. Solid type 2. Infiltrative type (ca. 33%)
5. Fibrolamellar type
• diffuse infiltrative (ca. 5%)
6. Spindle-cell type 3. Combination type (ca. 42%)
Degree of tumour differentiation
1. High differentiation2. Moderate differentiation 3. Low differentiation
4. No differentiation (anaplastic type)
Cytologic differentiation
1. Polygonal 2. Pleomorphic 3. Clear-cellular 4. Small-cellular
Tab. 37.5: Macroscopic forms, microscopic typing and degree of
tumour differentiation in hepatocellular carcinoma3.5.2 Microscopic tissue types
Dysplasia: This condition is defined as variably large, different-
shaped hepatocytes, mostly in localized groups, with enlarged, pleomorphic and hyperchromatic nuclei as well as enlarged nucleoli. These dysplastic hepatocytes may often be polyploid.Dysplasia can occur as a macrocellular (with eosinophilic cytoplasm) or microcellular (with basophilic cytoplasm and increased proliferation) variant. The latter is considered to be a precancerous stage.
• Dysplastic foci with a diameter of 1(⫺2) mm
consist of enriched, predominantly small-cellular dysplastic hepa- tocytes.• Dysplastic nodes with a diameter of 0.3⫺1.0 cm contain
an atypical architecture and cellular atypias; trabecular or pseudo- glandular structures are evident. Atypical hepatocytes are often clear-cellular, basophilic or steatotic. Fluent transition into HCC is seen occasionally. (87, 95)䉴 Histologically, six growth forms of HCC can be differentiated. (s.
tab. 37.5) The most common form is the
trabecular type, usually
comprising highly differentiated carcinomas with polygonal tumour cells similar to hepatocytes; they grow in multilayered tra- beculae and enclose blood spaces lined with endothelium (usually without Kupffer cells).•
Thepseudoglandular type is generally
found in combination with the trabecular form. It is characterized by the formation of gland-like structures containing detritus and bilic or liquid material.•
Thescirrhous type shows excessive depo-
sits of sclerosed connective tissue, which is relatively low in cells.The moderately differentiated tumour cells lie between the septa, which resemble connective tissue. This type is mostly found after chemotherapy or radiation therapy.
•
Thesolid type is an undif-
ferentiated HCC, with the tumour cells displaying considerable cel- lular polymorphism; the trabecular tissue pattern has disappeared.The tumour is compact due to compression of the sinusoids.
•
Differentiation is, however, only possible in rare cases, since there is often considerable heterogenicity within the tumour, i. e. dif- ferent tissue types may be found in the same HCC. (s. figs. 37.4, 37.5)
• Fibrolamellar HCC is rare; it consists of solid cell trabeculae
with connective-tissue septation and a capsule. (s. figs. 37.10, 37.11)• Spindle cell-like differentiated HCC is likewise a very rare
histological form with a fascicular-sarcomatous growth pattern.(111) Prognosis is significantly poorer than with other forms of HCC
(S. K akizoe et al., 1987)
.Fig. 37.4: Highly differentiated, macrotrabecular hepatocellular
carcinoma with hydropic swelling of the tumour cells (condition after alcohol injection)䉴 Histological tumour differentiation ranges from the highly dif-
ferentiated grade 1 to the undifferentiated grade 4. (s. tab. 37.5) It
does not provide sufficiently reliable information for giving a prog- nosis of HCC. This also applies to the above-mentioned histologi- cal growth forms⫺ with the exception of the fibrolamellar type, which generally has a more favourable prognosis.Fig. 37.5: Partly solid, partly pseudoacinar HCC with bile pigment
in canalicular structures3.5.3 Cytological differentiation
䉴 Polygonal cells are predominantly found in HCC. They have dense nuclei and lumpy, eosinophilic cytoplasm, which subse- quently becomes basophilic. Bile canaliculi are often observed.
•
Pleomorphous cells are present in immature tumours. They are
often smaller than the hepatocytes and may vary in their overall cell size as well as in the size of the nucleus; occasionally, they grow to form multinuclear cells or multinuclear giant cells contain- ing bizarre nuclei.•
Typical hyaline inclusion bodies with autofluor- escence are sometimes detected, pointing to disturbed protein secretion; they can be differentiated immunohistochemically, e. g.in the form of ferritin, α1-foetoprotein or α1-antitrypsin. The detection of Mallory bodies in tumour cells is likewise typical for HCC. They are caused by disturbed metabolism of the intermedi- ate filaments. Ground glass cells are also found (due to HBsAg expression or enhancement of fibrin). Nuclear inclusion bodies are caused by invaginated cytoplasmic substances. Bile may still be produced, particularly within a highly differentiated HCC; mor- phologically, this is expressed in the formation of bile thrombi.
•
The glycogen content in the tumour cells varies. When a large amount of glycogen (and water or lipids) is stored, HCC takes on a
clear-cellular “hypernephroid” form. • Small-cellular carcinomas
are also occasionally in evidence. (s. tab. 37.5)A
grading of HCC with regard to these different forms is usually
without clinical importance: (1.) the forms do not necessarily cor- relate with the clinical course and (2.) different forms can appear in one and the same patient.3.5.4 Metastatic spread
HCC metastasizes intrahepatically, haematogenically (50 ⫺60%) and lymphogenically (30%) into the regional lymph nodes, mainly (and also initially) below the dia- phragm. • Haematogenous metastases affect the lungs (40 ⫺50%) and skeletal system; metastases are only rar- ely found in the kidneys or brain. • However, infiltrating growth affects neighbouring organs (e. g. gall bladder, diaphragm, kidneys, hepatic vessels, bile ducts).
(78, 179)3.6 Symptomatology 3.6.1 Subjective complaints
HCC develops without subjective complaints. Even fur-
ther tumour growth often remains undetected; alterna-
tively, the complaints are explained as general symp- toms relating to cirrhosis or to an existing chronic liver disease. HCC is hence usually detected (too) late. Com- plaints are very varied:
pain in the upper abdomen weight loss bloating, flatulence fatigue, weakness inappetence, nausea stool irregularities The patient is suspected to be suffering from HCC when the subjective complaints continue to worsen and when an increase in complaints, which may occur quite abruptly in some cases, cannot be explained by the pro- gression of cirrhosis. Most patients are affected by pain radiating into the right side of the back or into the right shoulder and neck area. This is due to an irritation of the phrenic nerve caused by expansion of the tumour towards or even into the liver capsule. Anorexia is more pronounced. Occasionally, the course of HCC is acute, resembling liver failure or liver abscess.
3.6.2 Clinical findings
The clinical situation deteriorates rapidly: febrile tem- peratures and leucocytosis as well as subicterus are observed; there are also signs of encephalopathy. An arterial murmur can often be heard on auscultation, since the tumour is mainly supplied with blood from the hepatic artery. A sudden “blossoming” of vascular spi- der naevi is frequently seen with underlying cirrhosis.
Virchow’s lymphadenopathy may be present.
(84)Fever and leucocytosis frequently occur in tumour necrosis.
Clinical findings include:
fever arterial murmur
subicterus tenderness upon pressure meteorism/ascites palpable tumour
latent encephalopathy perihepatic friction Paraneoplastic findings: Occasionally, there are para- neoplastic symptoms or syndromes which vary greatly from individual to individual, such as erythrocytosis (due to enhanced expression of erythropoietin), poly- cythaemia, hyperparathyroidism with hypercalcaemia, hyperthyroidism, painful gynaecomastia
(14), osteo- arthropathy, pseudoporphyria, hypercholesterolaemia, hypertension
(16), polyneuropathy, polymyositis
(64), water diarrhoea syndrome, dermatomyositis, and an increase in the vitamin B
12-binding protein. Large and glycogen-rich tumours cause hypoglycaemia (possibly due to enhanced expression of IGF2).
3.7 Diagnosis
The efficacy of therapeutic measures, particularly the prospect of curative treatment and thus prognosis, depend upon the early diagnosis of HCC.
3.7.1 Laboratory findings
While non-specific signs of inflammation may also be present in liver cirrhosis, they are much more obvious in HCC. For example, positive CRP values as well as an increase in the BSR, α-globulins, α
1-antitrypsin and fibrinogen can be found. A constellation of decreased serum iron and elevated serum copper
(140)points to a consuming inflammatory process or malignant tumour.
(s. tab. 5.10) Usually, the haemogram shows leucocytosis and anaemia. Nitric oxide values in the plasma are ele- vated in correlation with tumour size. (s. tabs. 37.6, 37.11, 37.12)
1.
Non-specific parameters
• CRP ⫹, BSR 앖, α2-globulins앖, γ-globulins 앖 fibrinogen앖, cholinesterase 앗, D-dimer 앖 2.
Iron/copper constellation
• iron 앗, copper 앖, ferritin 앖 3.
Cholestasis
• γ-GT 앖, AP 앖 4.
Enzyme constellation
• GPT 앖, GOT 앖, LDH 앖, HBDH 앖, GDH 앖
5. • GOT/GPT ⫽ ⬎2
• γ-GT/GOT ⫽ ⬎12
• (GPT ⫹ GOT)/GDH ⫽ ⬍15 5.
Serology
• α1-foetoprotein앖 (⫹ ferritin 앖) ⫽ 앖앖
• des-γ-carboxy prothrombin 앖
• aldolase A 앖
• α-L-fucosidase 앖
Tab. 37.6: Laboratory findings arousing suspicion in HCC, with
increasing diagnostic reliability given by continuous deviation from normal valuesConspicuous enzyme activities include a disproportion- ate increase in γ-GT, AP (especially the Regan isoen- zyme, which is identical to the placenta isoenzyme) and LDH. This constellation is thought to be an important indicator of HCC, particularly when HBDH is also ele- vated. Intrahepatic cholestasis is invariably present in HCC. More and more foetal γ-GT is produced in car- cinomatous hepatocytes and is detectable at the biliary pole of the cells. Transaminases rise only slowly; the enzyme quotients may provide revealing information. (s.
tab. 5.6) • Hepatic synthesis capacity (albumin, cholines- terase, Quick’s value) decreases progressively. ChE is also greatly reduced in cases of malignant ascites. • The indocyanine green test is considered to be conclusive for evaluating the residual liver parenchyma prior to hepa- tectomy. A combination of the ICG test and galactose elimination capacity is an even more reliable tool, according to our experience.
␣
1-foetoprotein: AFP has gained great importance as a
laboratory value pointing to HCC. (s. p. 106) First dis-
covered in hepatomas by
G. Abelev et al.in animal experi-
ments in 1963, this type of oncofoetal protein was also
detected in human HCC by
Y. S. Tatarinovin 1964. AFP
is a glycoprotein formed initially within the yolk sac and later in the liver and gastrointestinal tract of the foetus.
Serum values of about 70,000 µg/l are found in neo- nates, decreasing to the normal value of < 10 µg/l within 9 ⫺12 months. Higher values can be detected in liver cell regeneration (acute and chronic hepatitis, cirrhosis) and particularly in HCC (> 20 µg/l). Thus, a continuous increase in AFP values arouses suspicion; a value of
> 100 µg/l is highly suspicious for HCC. There is only a moderate correlation between AFP and the respective tumour size and doubling time. A false-negative AFP value is detected in about 20% of patients suffering from HCC. The response to chemotherapy usually corres- ponds to decreasing AFP values. • Specificity is between 76 ⫺91%, sensitivity 39⫺64% (approx. 85% when ferri- tin is increased at the same time). Serum values of
> 2,000 µg/l may ultimately be reached. Liver metastases usually show values of < 150 µg/l, with no tendency to rise. AFP values in the normal range exclude HCC in 90 ⫺95% of cases. Generally, no AFP is produced in very mature or very immature tumour cells. AFP pro- duction in elder patients is lower; it is mostly higher in virus-related cirrhosis than in alcoholic cirrhosis.
(40, 75, 152, 154)Diagnostic importance is also attributed to des-γ-car- boxy prothrombin (60⫺80% positivity in HCC)
(H. A.Liebman et al., 1984)
.
(90, 120, 133, 154)It is synthesized in the normal hepatocytes and therefore also in HCC. The diagnostic accuracy in small hepatocellular carcinomas (< 3 cm) could be greatly improved by determining this precursor prothrombin (PIVKA II) in combination with AFP. • A decrease in the factor-II index, i.e. (factor VII
⫹ factor X) ⫺ (factor II) ⫽ >15, has proved to be a specific and independent marker of HCC. • In addition, isoferritins, Regan-AP, telomerase activity
(99)and L- fucosidase as well as CEA variants may be helpful in the demarcation of HCC. • Laboratory diagnosis of HCC is indeed much more reliable if various important param- eters are added. (s. tab. 37.6)
3.7.2 Imaging procedures
Sonography: Sonography is the method of choice in moni- toring the course of risk patients, particularly in combina- tion with AFP determination at (4 ⫺)6 month intervals.
In this way, a longer survival time could be achieved.
Under optimal examination conditions with an experi- enced investigator, foci are detectable at a size of 1 cm.
However, demarcation of HCC in a cirrhotic liver is very difficult, if not impossible, due to its inhomogen- eous reflex pattern. Small foci (< 2 cm) are usually hypo- echoic; when they grow, their echogenicity increases (due to deposition of fat, connective tissue formation or necrosis) and a hypoechoic margin with a rather blurred halo ( ⫽ tumour cells plus compressed liver parenchyma) is often in evidence. A bull’s eye is also detectable ( ⫽ hypoechoic centre surrounded by a hyperechoic edge).
At a size of > 4 cm, the reflex pattern becomes inhomo-
geneous. However, HCC has no typical image. In 50 to 70% of cases, diagnosis is successful at a size of 2 cm and more. Detection of HCC also depends on its degree of differentiation: well-differentiated liver carcinomas are barely distinguishable from normal hepatic tissue. A hypoechoic halo points to a fibrous capsule and ⫺ in combination with a mosaic structure within the focus ⫺ arouses suspicion of a carcinoma.
(2, 25, 34, 38, 40, 46, 87)• The use of colour-encoded Doppler sonography shows hypervascularization even in the early tumour stage and thus provides diagnostic differentiation from regenera- tion nodes and adenomatous hyperplasia. Visualization of tumour vessels is improved by intravenous injection of a contrast medium. When HCC infiltrates the portal vein, arterial vessels can be identified in the tumour thrombus in 60 ⫺70% of cases.
(8, 11, 56, 61, 71, 123)• The use of intraoperative US as well as laparoscopic US
(20,68)
results in a 20% higher detection rate of HCC than in preoperative diagnostics. When all space-occupying lesions of the liver are included, the diagnostic results are up to 40% better than those gained by conventional sonography. The haemodynamic pattern “nodule-in- nodule” ( ⫽ a vascular spot in a hypovascular nodule) is a sign of a suspicious node ⫺ this could, in turn, be an early stage of HCC.
(91, 181)• Characterization of liver tumours with contrast-enhanced sonography and digital grey-scale analysis offers the possibility of an investiga- tor-independent differential diagnosis.
(11)Computer tomography: The detection of small carcino-
mas, especially of tumours in cirrhosis, is unreliable
using conventional CT scanning. However, when the
tumour has reached a certain size, capsule formation,
fatty transformation of the tumour mass, vascular
infiltration and development of arterioportal shunts are
evident. Abdominal lymph node metastases can be
found. Sensitivity is about 60%. (s. fig. 37.6) • The intro-
duction of lipiodol CT has resulted in a considerable
improvement in HCC diagnosis: the oily contrast
medium injected into the hepatic artery under angio-
graphic monitoring is stored within the tumour tissue,
so that any tumour above a size of 3 mm can be iden-
tified in conventional CT scanning in full contrast for a
period of 1 ⫺2 weeks. Sensitivity is about 70%, specific-
ity about 80%.
(23, 113)• Spiral CT has proved to be
most effective (sensitivity 90%). This technique offers a
biphasic examination supported by contrast medium,
with the first (hepatic-arterial) phase showing mainly
hypervascularized HCC and the second (portal-venous)
phase showing mainly hypovascularized HCC.
(10, 37)•
CTAP (CT following arterioportography) is considered
to be a very sensitive method for the detection of HCC
in a non-cirrhotic liver (CTAP is not suitable in cases
of portal hypertension and portosystemic shunts). The
sensitivity of CTAP is 85 ⫺90%. • Although CT arterio-
graphy is an invasive procedure, it shows the highest sen-
sitivity in extensive cirrhosis, because HCC is predomi-
nantly supplied with arterial blood.
Fig. 37.6: HCC: nodular liver surface in cirrhosis. Tumour forma-
tion in the right lobe of liver, in parts at the margin; central hypo- density with peripheral hypervascularization (CT after CM)Angiography: HCC appears as a hypervascularized tumour in angiography via the hepatic artery. The ves- sels show irregular internal diameters; due to arteriove- nous anastomoses, the hepatic veins are rapidly filled, giving retrograde visualization of the portal vein. Poorly vascularized or non-vascularized areas present in the tumour develop due to necrosis or bleeding. Evidence of a hypervascularized “bush pattern” in a cirrhotic liver is considered to be an obvious sign of HCC. In terms of sensitivity, angiography is, however, inferior to other imaging techniques. There is no alternative to angiogra- phy when assessing surgical strategies or chemoperfu- sion and chemoembolization; angiography is often com- bined with simultaneous regional tumour treatment.
Magnetic resonance imaging: Malignant tumours gen- erally show longer relaxation times in comparison to healthy liver tissue. This results in hypointensity in T
1- weighting and hyperintensity in T
2-weighting ⫺ in con- trast to healthy liver parenchyma. T
2-signal intensity is more sensitive in the diagnosis of advanced HCC as well as in the assessment of a tumour capsule; this is prog- nostically significant. Early stages of highly differenti- ated HCC can be defined better in T
1-weighting. Diag- nosis is further improved by the use of contrast medium.
(28, 113, 170, 173)
Scintigraphy: The approaches used to date in scintiscan- ning have proved to be inadequate for diagnosing HCC and are also inferior to other imaging techniques. • New methods (Tc-GSA) as well as immunoscintigraphy,
18F- FDG PET or SPECT
(3, 86, 156, 160)have not yet been evaluated with regard to their diagnostic sensitivity or specificity in detecting HCC or in preoperative staging.
Their main value consists in better detection of extra- hepatic tumours; they may also have a direct impact on operative management.
(1)Plain X-ray: The rare event of HCC calcification is char- acterized by the so-called sunburst sign.
3.7.3 Morphological diagnosis
Percutaneous fine-needle biopsy: This technique is asso- ciated with the risk of tumour cell spreading. The fre- quency of subcutaneous implantation metastases is reported to be 2%; they generally appear within 3 months. When this procedure is indicated, the bleeding risk from the usually hypervascularized tumour must also be taken into account. The cytologic-diagnostic sensitivity is 80 ⫺85% of cases; specificity is 97⫺100%.
(31⫺33, 128)
Percutaneous thick-needle biopsy: We consider this tech- nique to be contraindicated, mainly because of the bleeding risk involved: (1.) HCC is usually a densely vascularized tumour, (2.) the tumour tissue does not possess the requisite spontaneous contractility for mechanical closure of the biopsy canal, (3.) local coagu- lation is usually disturbed in the tumour tissue. Punc- tures into the necrotic centre of a tumour may cause uncontrollable haemorrhagic oozing. The risk of implantation metastases is the same as in FNB.
(33)• This raises the question of whether histological or cyto- logical confirmation of the diagnosis is really necessary.
In this context, cytological assessment of the malig- nancy is equal or even superior to histology. However, cytological delimitation of highly differentiated HCC from adenomatous hyperplasia may be very difficult in certain cases. Diagnosing the tumour type is much more accurate by histological examination than by cytology.
Surgery is indicated if a resectable space-occupying lesion in liver cirrhosis with sufficient evidence of HCC has been determined by imaging techniques and the AFP value is > 400 µg/l. Operability in terms of internal medicine as well as hepatology has to be established. • Provided laboratory, sonographic and radiological find- ings are largely unambiguous, surgery is indicated even without an increase in AFP values; preoperative percu- taneous biopsy should not be carried out in such cases.
Morphological clarification of the findings is necessary if differential diagnosis of a hepatic space-occupying lesion is unclear. • Cytological or histological confirm- ation of tumour malignancy is likewise required prior to palliative therapy, even in patients with no chance of curative treatment.
Laparoscopy: This technique should always be used to achieve the required morphological clarification. (s. figs.
37.1 ⫺37.3, 37.7, 37.8, 37.16, 37.17)
(9, 88, 157)• It ini-
tially serves as explorative laparoscopy in the careful
inspection of the abdominal cavity using various body
positions in order to gain as reliable a picture as possible
of malignant metastases, particularly in the region of
diaphragm, peritoneum and ligaments. (s. figs. 37.27,
37.28) Inspecting large areas of the visible surface of the
liver as well as certain parts of the underside when lifted
by a probe allows excellent differentiation of all the
respective findings ⫺ especially in cirrhosis. Explorative
Fig. 37.7: Small to medium-nodular, alcoholic cirrhosis with undif-
ferentiated, multilobular hepatocellular carcinoma and subcapsu- lar vascularizationFig. 37.8: Large-bulbous hepatocellular carcinoma due to alcohol
abuse and active chronic hepatitis B with cirrhotic transformation in some placeslaparoscopy also facilitates staging, which cannot be achieved as effectively by imaging diagnostics.
All findings are documented by photolaparoscopy. (s. p.
156) Sufficiently large and compact tissue particles can be obtained from foci lying close to the surface by using biopsy forceps. Highly vascularized areas can thus be avoided. The site of extraction can be coagulated imme- diately, so that prolonged biopsy bleeding or late bleed- ing is almost completely excluded. • Foci that lie at a deeper level or those which cannot be reached with for- ceps should be evaluated by fine-needle biopsy. Assisted by sonography, this can be carried out at the same time as or after deflating the pneumoperitoneum, while the position of the trocar (or alternatively the Veres needle) is maintained. If necessary, a clear view into the abdo-
minal cavity can be regained by quick insufflation using the trocar.
Explorative laparotomy, as is occasionally recom- mended in the literature, is contraindicated in our opinion. • All comparisons with respect to the degree of risk, stress or inconvenience to the patient, expend- iture of time, material, staff and the costs involved, and particularly the diagnostic benefits, speak against laparotomy and in favour of explorative laparoscopy.
3.8 Prognosis
The prognosis of HCC is determined by the tumour mass and its speed of growth at the time of diagnosis.
Signs of a poor prognosis are included in the (p)TNM classification. (s. tab. 37.2)
1. Considerable size of tumour (> 5 cm) 2. Infiltrative or multilocular growth 3. Metastatic spread
The natural course of disease shows an average survival rate of 5 months (2 ⫺8 months). Longer courses of HCC have only been observed occasionally; some 3% of patients survive for 5 years. • The number of tumour nodes as well as the existing cirrhosis correlate with the likelihood of recurrence. In contrast, the histological degree of tumour differentiation as well as potential capsule or bile formation by the tumour are unlikely or still disputed risk factors for a relapse. This is also true of the biological features of the tumour, such as ploidy, mitosis rate and proliferation marker index.
(57, 108)By including clinical and laboratory parameters (ascites, jaundice, hyperalbuminaemia),
K. Okuda et al. (1985)introduced a classification into stages in order to improve prognostic accuracy regarding survival time.
(s. tab. 37.7) The survival rate in the untreated course was calculated to be 11 months in stage I, 3 months in stage II, and 1 month in stage III. The corresponding 1-year survival rates are 39%, 12%, and 3%. A survival rate of 29 months can be expected for tumours with
< 25% invasion of the liver in stage I; the corresponding figure for 25 ⫺50% liver invasion is 8 months. With smaller tumours (< 3 cm), the survival rates are slightly better: 1 year ⫽ 91%, 2 years ⫽ 55%, 3 years ⫽ 13%.
In older patients and in those with moderately increased γ-GT values, tumour growth is retarded. Evidence of antibodies against the tumour suppressor gene p53 point to an unfavourable prognosis.
Another alternative HCC staging system is the so-called
CLIP score
(35, 36). This also combines morphological
criteria of the HCC with liver functions (Child-Pugh)
and, additionally, with portal vein thrombosis as well as
AFP. (s. tab. 37.8) • BCLC classification can be recom-
0 points 1 point
A: Liver invasion ⱕ 50% ⬎ 50%
B: Ascites no yes
C: Bilirubin ⱕ 3 mg/dl ⬎ 3 g/dl
D: Albumin ⬎ 3 mg/dl ⬍ 3 g/dl
A⫹B⫹C⫹D ⫽ 0 points: stage I⫽ 8.3 A⫹B⫹C⫹D ⫽ 1⫺2 points: stage II⫽ 2.0 A⫹B⫹C⫹D ⫽ 3⫺4 points: stage III⫽ 0.7
Tab. 37.7: Classification of the stages of hepatocellular carcinoma
and survival time in months (untreated)(K. O kuda et al., 1985)
Criteria Points
0 1 2
1. Child-Pugh A B C
2. HCC morphology solitary multifocular multifocular
< 50% of liver < 50% of liver > 50% of liver 3.α1-fetoprotein < 400µg/l > 400µg/l
4. portal vein no yes
thrombosis
Survival rate
points mean rate 1 year (%) 2 year (%)
0 36 84 65
1 22 66 45
2 9 45 17
3 7 36 12
4⫺6 3 9 0
Tab. 37.8: CLIP staging system of HCC and the respective survival
rate (35, 36)mended for staging-adapted management of HCC.
(107, 108)
Complications: Acute liver failure, arterioportal fistula formation, oesophageal varices
(15)and pulmonary hypertension have been reported as complications. • In most cases, the cause of death is anorexia with tumour cachexia, accompanied by signs of circulatory and renal failure. Occasionally, there is intraperitoneal haemor- rhage, portal vein thrombosis
(138, 146)and tumour rup- ture with formation of haemorrhagic ascites.
(121)Spontaneous regression: As far as I know, it was
E. B.Gottfried et al. (1982)
who first reported spontaneous regression of HCC. In the meantime, further (> 25) un- usual observations of this kind have been published.
However, recurrence after such regression has also been reported.
(52, 62, 69, 76, 104, 112, 119, 162, 166)3.9 Therapy
Surgical, local-interventional, regional and systemic procedures and palliative measures are available for the treatment of HCC. • Longer survival periods can be achieved by surgery, which is considered to be the only form of potentially curative treatment for small tumours (< 3 cm, or even < 5 cm).
At present, merely 5% of patients suffering from HCC have any chance of being cured. Some 25 ⫺30% of inop- erable patients were rendered operable by means of pre- operative selective irradiation with yttrium
90micropar- ticles (intra-arterially) or by several cycles of cisplatin ⫹ IFN- α
2b⫹ doxorubicin ⫹ 5-fluorouracil. The results could be further improved by postoperative treatment with lipiodol-J
131. • In this context, the overall chance of recovery is reported to be 15%.
䉴 An interdisciplinary consensus conference to decide upon the best therapy possible must be held before any treatment of a carcinoma begins. A gastroenter- ologist, a surgeon, an oncologist and a radiologist should participate, irrespective of the department in which the patient is being cared for. The treatment plan agreed upon must then be explained to the patient and (with the patient’s consent) also to the closest relatives. • So-called “virtual liver surgery” will play an important role in this connection in the fore- seeable future.
3.9.1 Surgical therapy Resection
Only about 30% of tumours are still resectable by the time HCC is diagnosed. • The following are considered to be contraindications for resection: (1.) multicentric tumours, (2.) metastases, (3.) invasion of the tumour into the portal or hepatic vein, (4.) liver malfunction (jaundice, hypoalbuminaemia, considerable reduction in Quick’s value and cholinesterase), and (5.) decompen- sated portal hypertension (ascites, encephalopathy). • Surgical techniques include: (1.) right-sided or left-sided hemihepatectomy, (2.) right-sided or left-sided lobec- tomy, (3.) segment IVb resection, and (4.) atypical resec- tions and wedge excisions. (s. pp 800, 870) Resection of the small left lobe of liver is tolerated best. When the larger right lobe is resected, there is a risk of liver insuf- ficiency due to inadequate residual functional capacity.
Preoperative administration of
131J-lipiodol (60 mCi/
injection) showed promising results.
(143)• The follow- ing procedure was hence recommended for the safe removal of liver parenchyma: first, embolization of the portal vein branch associated with the tumour is carried out; this results in atrophy of the embolized lobe and hypertrophy of those segments to be maintained, with a correspondingly better functional capacity. • If non- resectability is revealed during the operation, a catheter should immediately be implanted in the hepatic artery for regional chemotherapy. • For patients without cir- rhosis, mortality is 1⫺3%, for those with cirrhosis 7 ⫺25%. The survival rate is 25⫺60% after 5 years, depending on tumour size; this rate sinks to 25 ⫺35%
when the capsule or the portal vein are infiltrated. De-
cisive factors are the radicalness of the intervention and
the functional capacity of the residual liver parenchyma.
Intraoperative blood loss should be kept as low as pos- sible in order to limit morbidity. Patients in Child-Pugh stage A showed a 5-year survival rate of 27 ⫺53% (cf.
disease-free survival rate of 20 ⫺33%), and in stage B of about 30%; stage C is usually non-resectable or has no surgical benefits. Preoperative staging with the help of laparoscopy is absolutely necessary. It is decisive for choosing the correct kind of management (the original procedure had to be changed in 20 ⫺30% of patients due to the laparoscopic findings).
(9, 88) (7, 17, 63, 80, 96, 98, 117, 125, 141, 182)(s. p. 799)
The radicalness of the intervention has to be assessed by the pathologist on the basis of the findings of the resected preparation as well as the surgical report. First, the pTNM stage is determined. (s. tab. 37.2) Then, the so-called R classification is carried out in order to estab- lish whether the resection was curative. R0 resection means that the tumour was completely removed, no tumourous residues were left behind, and the resection margin is tumour-free. R1 resection implies that there are still tumour cells at the resection margin. R2 resec- tion signifies that macroscopically visible tumour tissue is left. When the tumour markers normalize within the first 4 postoperative months after R0 resection, this is defined as R0a resection, and without normalization as R0b resection.
(7)• In cirrhosis patients with small HCC, liver function is the most important factor for long-term survival. In this context, the indocyanine green test, MEGX test and the Okuda stage system have proved useful. Under favourable conditions, relapse resections are also possible. • Resection therapy is sometimes used as “bridging” prior to a liver transplantation.
Recidivation is generally due to undetected small intra- hepatic foci. Especially in cirrhosis, the recurrence rate is very high as a result of this factor. A further cause of recidivation after primary R0 resection is attributed to the multicentricity of the HCC, i. e. synchronic or meta- chronic development of additional tumours which are independent of the primary tumour. Therefore, the secondary prophylaxis takes on a special meaning (sys- temic or intra-arterial chemotherapy, interferon, reti- noids, autologous lymphocyte transfusion, etc.). An increase in ornithine decarboxylase and spermidine is apparently a high risk factor for recurrence. • Even after recidivation, repeated resection or interventional pro- cedures (e. g. RFTA) are sometimes successful.
Cryosurgery: Cryotherapy of tumours was introduced by
I. S. Cooperin 1963. • Intraoperative insertion of a probe into the liver tumour is carried out directly.
Necrosis is induced in the tumour tissue by the applica- tion of fluid nitrogen as well as by the subsequent rethawing of the iced area (ca. ⫺50°C). This procedure is usually repeated. It is indicated in cases of non-resect- able liver tumours and for the freezing of margins following resection.
(21, 175)(s. p. 798)
Liver transplantation
With insufficient functional reserve capacity of the liver and/or non-curative resectability of the tumour, liver transplantation is generally indicated. However, such a procedure does not yield better long-term results than resection, particularly since the frequency of relapse (some 65% of cases) in the transplanted liver is higher due to the fact that immunosuppression is constantly required. A further problem is that the tumour-doubling time (102 ⫺195 days) is shortened to about 26 days.
Relapses originate from preoperatively undetected metastases, which show strikingly rapid growth. • In order to limit these relapses, a method called neoadju- vant chemotherapy has been recommended, e.g. intra- arterial or systemic administration of cisplatin and dox- orubicin, possibly in combination with interferon. The survival rates after 1, 2 and 3 years were reported as 70%, 60%, and 59%, respectively. Another form of neo- adjuvant therapy was based on 5-FU (continuous infu- sion for 6 months), with intermittent administration of adriamycin and cisplatin prior to transplantation. This procedure showed a survival rate after 1, 2 and 4 years of 73% vs 55%, 61% vs 40%, and 61% vs 22%, respec- tively. The relapse rate was only 18%. • Indications for liver transplantation should always be considered with great caution. Careful preoperative examination regard- ing HBV or HCV infection (PCR, if necessary) is recom- mended, since there is a very high risk of the trans- planted liver being infected by viruses from extrahepatic tissues under immunosuppression. • Survival rates in HCC cases without and with cirrhosis are almost equal at up to 85% after 1 year, 30% after 3 years, 20 ⫺45%
after 5 years, and about 20% after 10 years. A long-term survival rate is only to be expected in patients with small solitary tumour nodes (< 5 cm) or not more than 3 nodes of < 3 cm in diameter and without signs of vascu- lar invasion (so-called Milano or Mazzaferro criteria).
(114)
A hyperextended indication is only given in rare cases. These rates largely correspond to the results obtained in non-malignant liver diseases. Transplanta- tion is regarded as being better than resection in cirrho- sis patients with early detected HCC, since the precan- cerous potential of cirrhosis is totally eliminated at an early stage. In every case, preoperative and/or postoper- ative chemotherapy should always be considered. The previous operative mortality rate of 15⫺20% has now been further reduced.
(22, 56, 88, 117, 122, 144, 153, 158)3.9.2 Interventional therapy
After discussing the clinical and imaging findings as
well as the result of exploratory laparoscopy, it is
important to decide on the most promising management
in the interdisciplinary conference. • Once resection
techniques and liver transplantation have been ruled
out, the next alternative are interventional procedures
(s. tab. 37.9).
1. Percutaneous injection therapy
앫 ethanol앫 acetic acid
2. Transcatheter arterial therapy
앫 transarterial chemotherapy 앫 transarterial embolization 앫 transarterial chemoembolization 앫 transarterial radiation⫺ 131J-lipiodol
⫺ 90yttrium
3. Percutaneous thermoablation therapy
앫 laser-induced thermotherapy 앫 microwave coagulation 앫 radiofrequency ablationTab. 37.9: Local-interventional procedures in the management of
HCCPercutaneous injection therapy
Percutaneous ethanol injection (PEI): This procedure was described by
N. Sugiura et al.in 1983. In solitary HCC with a size of < 3 ( ⫺5) cm in diameter, alcohol (96 vol.%) is injected into the tumour using sonographic, laparoscopic
(79)or CT monitoring. This technique (5 ⫺10 ml, about 3 times a week, regularly over a period of 3 ⫺4 months) is a relatively inexpensive procedure which is technically simple and well tolerated (albeit usually very painful!). Due to the fact that HCC has a soft consistency compared to the firmer liver tissue, it is possible to a large extent to infiltrate the tumour selec- tively and limit necrosis as desired. (s. fig. 37.9)
Fig. 37.9: Subcapsular HCC in liver cirrhosis: mostly necrotic
following ethanol injectionImaging findings after PEI are: in US, alcohol distribu- tion appears like an echogenic cloud and the tumour becomes relatively hyperechoic; in Doppler sonography, the colour signals disappear; in CT, gas formation is sometimes visible. More than 70% of the tumour mass may be necrotized. Under favourable conditions, a com- plete response can be achieved in 80% of cases. • Adverse events include acute cholecystitis, cholangitis, haemorrhages, liver abscess, pleural effusion, portal vein thrombosis, etc. • The PEI technique has no significant
effect on tumours with a size of > 5 cm and multilocular tumours. Some cases of stitch-track metastases have been reported. The relapse rate is approx. 60%. The sur- vival rate is 68 ⫺80% after 3 years and about 50% after 5 years. Long-term success depends upon the number and size of the tumours as well as on the respective liver function. PEI is contraindicated in Child-Pugh stage C.
(18, 72, 79, 85, 89, 105, 115, 118, 134, 139, 177)
Percutaneous acetic acid injection (PAI): Injection of acetic acid (40 ⫺50%, 2⫺5 ml, 2⫺4 sessions) into the tumour also resulted in necrotic destruction of the tumour tissue
(M. Imamura et al., 1995). PAI is considered to be just as effective as PEI; the success was most evi- dent when the intratumoural retention of acetic acid persisted for about three days after completion of the treatment.
(66, 102, 132)Transcatheter arterial therapy
As an alternative to percutaneous injection therapy, there is a possibility to develop new local-interventional treatment strategies by selectively exploring the tumour- feeding branch of the proper hepatic artery.
Transarterial chemotherapy (TAC): The systemic admin- istration of cytostatics, both as mono- or polychemo- therapy, led to unsatisfactory results accompanied by considerable side effects. By transporting cytostatics transarterially to the tumour itself, it was possible to build up a high concentration of these substances in the arteries supplying the HCC. Due to the fact that there were far fewer side effects, the dosage of cytostatics could be optimized. A further advantage of this method is that the extratumoural parenchyma has a portal venous blood supply; thus the arterially transported cytostatics do not cause any significant peritumoural liver damage. This transarterial chemoperfusion could be improved even more by positioning the selective-arte- rial catheter exactly and by implanting a port system.
(43)
