Clinical Research Med Sci Monit, 2000; 6(5): 915-928

Causes and pathomechanisms of oesophageal varices development

Karl-Joseph Paquet

Department of Surgery and Justice, Medical Service of Lower Saxonia, Hannover, Germany

key words: pathomechanism, pathophysiological background, development, treatment, oesophagogastric varices

SUMMARY

Portal hypertension is a common clinical syndrome with chronic liver diseases and is characterised by a patho-
logical increase in portal pressure. Moreover, portal hypertension is associated with increased portal blood flow.
Increased vascular resistance in portal hypertension is because of an increase in both intrahepatic and portosys-
temic collateral resistance. Chronic elevations in systemic and splanchnic blood flow have been documented as
key elements of hyperdynamic circulatory state of hypertensive animals and humans. Peripheral vasodilatation
initiates the development of the classic profile of decreased systemic elevated splanchnic blood flow and elevat-
ed cardia index that characterises this state.
Portosystemic collaterals develop as a result of portal hypertension. This is the central pathophysiological event
that leads to bleeding from oesophagogastric varices and portosystemic encephalopathy. Collateral vessels
respond to various vasoconstrictors and vasodilators.- Varices in the distal 5 cm of the distal oesophagus are
easily identified by endoscopy because of their superficial location in the lamina propria and therefore are must
apt to bleed and why the current practise of endoscopic therapy is likely to be successful in obliterating the
varices. In patients with oesophageal varices the dilated deep intrinsic veins displace the superficial venous
plexus, assume a supepitheal position and are endoscopically visible as teleangiectasia, cherry red spots, red
colour signs, hemocystic spots, red wale markings or varices on varices. As alternative endoscopic way of treat-
ment the paravariceal injection has been propagated by our group thus preserving the pathophysiologic collater-
als and preventing early new formation of collaterals and rebleeding.
Pathophysiologically the concept of erosion has been abandoned and replaced by the explosion theory: bleeding
probably occurs when the expanding force by pressure and flow can no longer be counter-balanced by the
variceal wall tension; at this point the varices rupture and bleed. When the varix distension has increased, the
radius has increased and the wall thickness decreased. Thus early diagnosis of patients with a high tendency to
bleed can easily be made by endoscopy, measuring portal and / or oesophageal-variceal pressure and character-
istising the chronic liver disease according to the Child-Pugh-classification.

INTRODUCTION quences of portal hypertension are gastrointestinal

bleeding from ruptured oesophagogastric varices,
Portal hypertension is a common clinical syndrome ascites formation and hepatic encephalopathy.
with chronic liver diseases and is characterised by a
pathological increase in portal pressure. Elevated A grasp of the biological mechanisms involved in
portal pressure results in extensive formation of the pathogenesis of portal hypertension is essential
porto-systemic collaterals that divert portal blood to an understanding of the complications of chro-
to the systemic circulation. This causes many nic liver disease and to the development of rational
haemodynamic disturbances, which result in com- therapies. Thus, an overview of the basic patho-
plications of portal hypertension. The major conse- physiological mechanisms of the splanchnic and

Received: 1999.06.01 Correspondence address: Karl-Joseph Paquet, MD FACS FICS, Department of Surgery and Justice, Medical Service of Lower Saxonia,
Accepted: 2000.07.08 Hildesheimer Str. 41, D-30169 Hannover, Germany

915
Clinical Research

Increased resistance

Portal hypertension is associated with increased

resistance to portal blood flow. Increased vascular
resistance is because of an increase in both intra-
hepatic and portosystemic collateral resistance in
comparison with the low resistance to the normal
liver.

Figure 1. Contributing factors to portal hypertension Intrahepatic resistance

The normal liver is a very compliant organ. Hence,

systemic circulatory derangements that lead to por- the intrahepatic resistance decrease with increase
tal hypertension and cause portosystemic collater- in blood flow because of distension of the vascular
als to develop is given. tree in response to increased inflow. This compen-
satory mechanism maintains the portal pressure
Some recapitulation of general principles is essen- within normal limits with a wide range of portal
tial to the understanding of the pathophysiology of flow in normal liver (Greenway and Stark, 1971)
portal hypertension. When OHM’s law is applied [1]. This phenomenon is not seen in the portal
to the vascular system, the pressure gradient hypertensive states in which the intrahepatic resis-
between two points (P1-P2) in a blood vessel can tance becomes fixed due to fibrosis and mechani-
be described as the product of blood flow (Q) and cal distortion of the vascular tree and the hepatic
resistance (R). vascular compliance is greatly reduced. Moreover,
the splanchnic flow is increased as discussed in the
P1 - P2=Q x R subsequent sections. Decreased compliance and
increased blood flow are probably instrumental in
Unlike pressure and flow, resistance can not be initiating and perpetuating the portal hypertension.
directly measured, but it can be derived from pres-
sure and flow. However, resistance to the flow of As stated above, there is little resistance in the nor-
blood in a vessel is best understood when mal liver, and the portal pressure remains low
expressed according to Pouseuille’s law: (4–8 mmHG) over a wide range of portal flows.
The main site of resistance in normal livers is some-
8hL what controversial. The hepatic sinusoids (where
R = ———
pr4 stellate cells are present), terminal hepatic venules
and also portal venules have been suggested as
in which h=coefficient of viscosity, L=length of possible sites of resistance. However, in view of the
vessel and r=radius of vessel. Expressed in these minimal contribution of the intrahepatic resistance
terms, substitution of resistance (R)=into Ohm’s to portal pressure in the normal liver, this issue is of
equation yields little importance. The flow into the portal system is
actively regulated by changes in vascular resistance
8hL at the level of splanchnic arterioles and not by the
P1 - P2 = Q ———
pr4 liver itself (Greenway and Stark, 1971) [1]. In portal
hypertensive syndromes, increased resistance to
Under physiological conditions, the length of blood portal venous flow may be localized to pre-hepa-
vessel (L) can be assumed to be constant. Similarly, tic, post-hepatic, or intrahepatic (pre-sinusoidal,
unless there are large changes in hematocrit, the sinusoidal or post-sinusoidal) sites (Genecin and
viscosity (h) is taken as constant. Hence, the Groszman) 1993) [2] (Figure 2). In pre-hepatic and
changes in pressure (P1-P2) are largely accounted post-hepatic portal hypertension, increased resis-
for the changes in flow (Q) and inversely related to tance is secondary to obstruction of portal venous
the changes in the radius (r of the vessel). The con- inflow or hepatic venous outflow, respectively.
tribution of these factors to the portal hypertensive Unlike pre- and post-hepatic hypertension, the
syndrome (Figure 1) will be discussed in the subse- intrahepatic sydromes are more complex and
quent sections. rarely can be classified according to a single site of
resistance.

916
 Paquet KJ – Causes and pathomechanisms of oesophageal varices…

example, in hepatic schistosomiasis, the increased

intrahepatic resistance results from granulomas
located in the presinusoidal areas (Beker and
Valencia-Parparcen, 1968) [6]. However, in late
stages, an elevated hepatic wedge venous pressure
gradient may be observed, reflecting increased
sinusoidal resistance. Chronic hepatitis seems to
have both pre-sinusoidal and sinusoidal abnormali-
ties, which increasingly contribute to vascular resis-
tance as the lesion progresses towards cirrhosis
(van Leeuwen et al, 1991) [7]. In ethanol-induced
liver disease, the resistance is increased because of
lesions in sinusoidal and post-sinusoidal sites. The
terminal hepatic vein fibrosis (or sclerosis),
encroachment on sinusoids by enlarged hepato-
cytes, collagen deposition in the presinusoidal
region or the space of Dissé, result in elevated por-
tal pressure even in the pre-cirrhotic stage
(Schaffner and Popper, 1963; Reynolds et al,
1969; Lieber et al, 1976) [8–10]. However, in
advanced stage of cirrhosis, regenerating nodules
Figure 2. Sites of obstruction (resistance) to portal venous flow and and pruning of the vascular tree contribute to
measurement of portal pressure. The schematic illustrates increased vascular resistance. A variety of other
the major locations of extrahepatic (prehepatic and posthep- non-alcoholic liver diseases cause portal hyperten-
atic) and intrahepatic (presinusoidal, sinusoidal, and postsi- sion because of increased sinusoidal vascular resi-
nusoidal) obstruction. A catheter tip is also shown wedged stance. In hepatic amyloidosis, resistance is
into a small hepatic vein for measurement of the WHVP. increased due to deposition of amyloid in the
When the catheter is withdrawn into the hepatic vein, FHVP space of Dissé (Brion et al, 1991) [11]. The capillar-
is obtained. HVPG = WHVP – FHVP. Direct measurement of ization process and occlusion of the fenestrae are
the PVP is accomplished intraoperatively either by catheteri- postulated to increase the resistance of passage of
zation of the umbilical vein or the portal vein by the trans- fluid across the endothelium, but the extent of
jugular or -hepatic approach. HVPG = hepatic venous pres- their effect on resistance to blood flow is unknown.
sure gradient; WHVP = wedged hepatic venous pressure,
FHVP = free hepatic venous pressure: PVP = portal venous The morphological changes that occur in chronic
pressure. liver diseases are undoubtedly the most important
factor involved in the increased intrahepatic resi-
stance. However, recent data also suggest a role of
An early view of vascular resistance in cirrhotic li- functional factors that lead to increased vascular
vers hypothesized that portal hypertension is the tone, similar to that which is seen in the arterial
consequence of a vascular obliterative process with hypertension. In chronic liver disease and also dur-
scar tissue and regenerative nodules, both occlud- ing acute liver injury, hepatic stellate cells acquire
ing and compressing vascular structures (Popper, contractile properties and may contribute to the
1958; Baldus and Hoffbauer, 1963) [3,4]. Thus dynamic modulation of intrahepatic resistance
earlier understanding of intrahepatic portal hyper- (Pinzani et al, 1992) [12]. These cells may act as
tension emphasized the role of anatomical alter- pericytes, a type of cell has been shown to regulate
ations leading to mechanical obstruction (irre- blood flow in other organs. The hepatic stellate
versible component) in the increased intrahepatic cells, which are also the main source of collagen
resistance. However, recently it has been shown synthesis, may contribute to the regulation of
that there is also an increased vascular tone in the hepatic flow at the microcirculatory level. Stellate
cirrhotic liver (reversible component) (Bhathal, cells are strategically located in the sinusoids with
1985) [5]. perisinusoidal and intrahepatocellular branching
processes that contain actin-like filaments. They
In intrahepatic portal hypertension, there may be also express the alpha smooth-muscle actin gene,
several areas of obstruction, and as the disease which is characteristic of vascular smooth muscle
progresses, new sites may become involved. For cells. The characteristics of these cells make them

917
Clinical Research

similar to myofibroblast. Myofibroblast-like cells Portosystemic collateral resistance

have been shown to exist in fibrous septa around
sinusoids and terminal hepatic venules in cirrhotic Although the collateral circulation begins as a con-
livers (Rudolph et al, 1979) [13]. These cells are sequence of portal hypertension, it evolves into an
postulated to play a role in the regulation of vascu- important mediator of the circulatory derangement
lar resistance in the cirrhotic rat liver (Bhathal and of portal hypertension in its own right. The por-
Groszman, 1985) [5]. tosystemic collaterals provide a route to decom-
press the hypertensive portal systems; however,
The vascular endothelium synthesizes vasodilators the vascular resistance of this collateral bed is still
such as nitric oxide, prostacyclins, hyperpolarizing greater than the resistance of the normal liver.
factor and vasoconstrictors such as endothelins Hence, portosystemic collaterals do not permit a
and prostanoids (Rubnayi, 1990; Vane et al, 1990) complete decompression. Development of por-
[14,15]. These vasoactive substances act in a tosystemic collaterals is discussed in detail later.
paracrine fashion on the underlying vascular
smooth muscle and modulate vascular tone. HYPERDYNAMIC CIRCULATION IN PORTAL
Normal vascular tone is maintained by a delicate HYPERTENSION
balance between these vasodilatory and vasocon-
strictive substances. Perturbation of this balance Chronic elevations in systemic and splanchnic
leads to abnormal vascular tone. Increased vascu- blood flow have been documented as key ele-
lar tone seen in cirrhotic livers could be because ments of hyperdynamic circulatory state (HCS) of
of a deficit of endothelial vasodilators or an portal hypertensive animals and humans. Periphe-
increase in the vasoconstrictors, or combination of ral vasodilatation initiates the development of the
both. Nitric oxide is a potent endothelial vasodila- classic profile of decreased systemic elevated
tor that has been shown to play an important role splanchnic blood flow; and elevated cardiac index
in the modulation of intrahepatic vascular tone in that characterizes this state (Colombato et al, 1991)
normal livers (Mittal et al, 1994) [16]. Preliminary [20].
evidence from laboratory investigations suggest
that there may be deficit of nitric oxide in the cir- Systemic vasodilatation
rhotic intrahepatic circulation (Gupta et al, 1995)
[17]. Using an isolated rat liver perfusion model, At least three mechanisms are thought to con-
this group has demonstrated that there is endothe- tribute to this peripheral vasodilatation: (i)
lial dysfunction in the intraheptatic microcircula- increased concentrations of circulating vasodila-
tion of cirrhotic livers (Gupta and Groszman, tors; (ii) increased endothelial production of local
1995) [18]. Endothelial dysfunction leads to vasodilators; and (iii) decreased vascular respon-
impaired release of endothelial vasodilators which, siveness to endogenous vasoconstrictors. The last
in part, may be responsible for the increased vas- mechanism is probably because of the effect of the
cular tone observed in cirrhotic livers. More first two components. The relative importance of
recently, it has been shown that the stellate cells each of these potential causes of peripheral vasodi-
(myofibroblasts) from cirrhotic livers exhibit an latation is unknown. Unfortunately I have no time
enhanced response to endothelins (Rockey and to describe in detail the acting substances like cir-
Weisiger, 1996) [19]. An imbalance between culating bile acids and glucagon as circulating
endothelial vasodilators and vasoconstrictors can vasodilators or the increased endothelial produc-
affect the activated stellate cells (myofibroblasts), tion of vasodilator like nitric oxide and
which modulate intrahepatic vascular tone. It is prostaglandins or the decreased response to vaso-
possible that both a deficit of vasodilators and an constrictors, the so-called hyporeactivity of the
increase in vasoconstrictors may be responsible for endogenous vasoconstrictors norepinephrine and
the increased vascular tone. vasopressin nor can I describe in detail the role of
the plasma volume expanding in portal hyperten-
In summary, there are multiple factors that may sion which is known since many years (Liebermann
lead to increased resistance to portal blood flow. and Reynolds, 1967) [21].
Some of these are irreversible, such as fibrosis, cap-
illarization, regenerating nodules, and some are The studies that examine the role of vasodilatation
quite dynamic such as the imbalance between and plasma volume expansion in the hyperdyna-
endothelial factors that leads to increased vascular mic circulation provide support for the peripheral
tone. vasodilatation hypothesis. According to this hypo-

918
 Paquet KJ – Causes and pathomechanisms of oesophageal varices…

have been shown to ameliorate the development of

the collateralization of the portal system thus impli-
cating increased portal pressure in the pathophysi-
ology of collateral formation (Halvorsen and
Myking, 1974) [22]. Lee et al (1993) however,
demonstrated that collateral formation could be
ameliorated by preventing an increase in splanchnic
blood flow with any decrease in portal pressure.
Thus, increased portal pressure does not appear to
be an absolutely necessary component of collateral
development in portal hypertensive states. There-
fore, propanolol and clonidine may prevent colla-
teral formation, in part at least, by decreasing flow,
not by their effect on portal pressure. On the other
hand, Sikuler et at (1985) [23] studied partial portal
vein ligation (PVL) rats on postoperative day 3 and
found no relationship between portosystemic
shunting and portal venous inflow. Evidence from
studies conducted in PVL rats suggest that increased
nitric oxide (NO) may be responsible for the initial
collateralization of the portal system. Increased NO
could result from progressive increases in spanchnic
flow, a potent stimulus known to be present in por-
tal hypertensive states. In addition, development of
a new portocollateral bed renders available a new
endothelial surface capable of producing NO (Lee
Figure 3. Hyperdynamic circulatory state in portal hypertension et al, 1993) [24,25].

Mosca et al (1992) [25] determined the vascular

thesis, portal hypertension leads to a relative hypo- responsiveness of collateral vessels to various vaso-
volaemia induced by the dilatation of the systemic constrictors and vasodilators using an in situ per-
and splanchnic circulation. This results in underfill- fused animal model. By administering norepineph-
ing of the systemic vascular space with consequent rine, 5-hydroxytryptamine, isoproterenol and acetyl-
decrease in central blood volume. This in turn choline in the presence and absence of their respec-
leads to activation of the sympathetic nervous sys- tive blockers, phentolamine, propanolol, and
tem, renin-angiostension system and release of L-NNA, they determined that collateral veins appear
antidiuretic hormone. Mediators from these sys- to be as sensitive to NO as arteries (Table 1).
tems result in sodium and water retention by the
kidneys, which results in increased plasma volume.
Peripheral vasodilatation and increased plasma vol- Table 1. Vasoactive mediators in portal hypertension.
ume result in a hyperdynamic circulatory state
(Figure 3). Vasodilators Vasoconstrictors

DEVELOPMENT OF PORTOSYSTEMIC COLLATER- Glucagon Norepinephrine

ALS AND OESOPHAGEAL VARICES Prostacyclin Serotonin
Substance P Endothelins
Portosystemic collaterals develop as a result of por- Atrial natriuretic factor Angiotensin II
tal hypertension. The development of collaterals is Bile acids Vasopressin
the central pathophysiological event that leads to Histamine
variceal bleeding and portosystemic encephalopa- Vasoactive intestinal peptide
thy in patients with portal hypertension. g -Aminobutyric acid
Leu- and Meta-encephalins
The exact nature of the physiological stimuli Endotoxin
responsible for initiating the collateral formation Tumor necrosis factor -a
remains controversial. Propanolol and clonidine Nitric oxide

919
Clinical Research

Regardless of the specific initiating stimulus, at least

at the beginning, the mechanism of collateral for-
mation appears to be recruitment of performed
channels, as opposed to de novo formation of new
vessels. Portosystemic shunting can be detected
almost immediately after the induction of portal
hypertension, a pattern of timing that is consistent
with the rapid dilatation of preformed vessels
(Mosca et al, 1992) [26]. Whether these preformed
vessels expand as the result to increased portal
pressure is unclear. Neoformation of vessels cannot
be excluded.

In humans, the portosystemic shunting of blood

occurs between the short gastric coronary veins,
and the oesophageal – azygos and the intercostal
veins; superior and the middle and inferior haem-
orrhoidal veins; the paraumbilicus venous plexus
and the venous system of abdominal organs juxta-
posed with the retroperitoneum and abdominal
wall. As discussed earlier, the dilatation of the pre-
existing embryonic channels is thought to be the
main mechanism of evolution of these collaterals.
With progression of portal hypertension, the num-
ber and diameter of these collaterals increases.

The development of gastro-oesophageal varices

constitute a major complication of portal hyperten-
sion. A threshold portal pressure gradient (HPVG) Figure 4. Radiograph of the venous zones in the human gastrooe-
of 11 to 12 mm HG has been shown to exist in sophageal junction (GZ, gastric zone; PZ, palisade zone; TZ,
humans with oesophageal varices, below which truncal zone) from Vianna et al 1987 (32)
varices are not encountered (Garcia-Tsao et al,
1985) [27]. Elevation above this pressure, however,
does not always result in varices, since many fied by endoscopy because of their superficial loca-
patients with HPVG in excess of 12 mmHG do not tion in the lamina propria [28]. Gastric varices, on
have varices. the other hand, are located in the submucosa and
are not easily visualised [29]. Anatomical studies of
Anatomic considerations the lower oesophagus have provided insight why
varices in the distal 5 cm of the oesophagus are
The portal circulation in normal individuals is a most apt to bleed and why the current practise of
high-flow (1000–1200 ml/min.), low pressure sys- endoscopic therapy is likely to be successful in
tem (7 mmHG) and provide the liver with 60% to obliterating the varices. Our current understanding
70% of its total oxygen supply. Both intrahepatic of the venous anatomy of this region derives from
and extrahepatic abnormalities that increase the the pioneering work of BUTLER in 1951 [30] and
resistance to blood flow in this system result in por- from recent studies which have used more sophis-
tal hypertension. Once portal hypertension devel- ticated, high-resolution techniques [31–36].
ops, extrahepatic collateral venous channels
between the portal and the systemic circulation In normal patients, four distinct venous zones have
dilate and form varices. been identified in the distal oesophagus and proxi-
mal stomach (Figure 4). Of these, the palisade
Varices can be found along the entire gastrointesti- zone and the perforating zone are considered the
nal tract in patients with portal hypertension. most important in the development and rupture of
Those most likely to bleed are found in the distal oesophageal varices. The palisade zone begins at
oesophagus and proximal stomach. The varices in the gastrooesophageal junction and extends 2–3
the distal 5 cm of the oesophagus are easily identi- cm cephalad. The veins within this zone are longi-

920
 Paquet KJ – Causes and pathomechanisms of oesophageal varices…

3–5 cm proximal to the gastrooesophageal junc-

tion. In the perforating zone, four distinct layers of
veins have been identified (Figure 5). Fine, superfi-
cial, arborizing intrepithelial veins drain capillary
beds in the epithelium into a superficial venous
plexus. Blood subsequently flows into three to five,
larger calibre, deep intrinsic veins. Blood from
these intraoesophageal veins is directed into peri-
oesophageal veins through perforating veins which
Figure 5. The layers of veins at the human distal oesophagus (From have valves to encourage bi-directional flow.
Kitano et al 1986 (31))
These normal vascular relations are altered signifi-
cantly in the patient with portal hypertension.
tudinally arranged in the lamina propria. Blood Dilatation and high pressures in the perioe-
flow within the palisade zone is bi-directional and sophageal veins cause the valves in the perforating
accommodates the variable pressures at the gas- veins to become incompetent, allowing retrograde
trooesophageal junction associated with the respi- flow into the deep intrinsic veins and its tributaries.
ratory cycle, coughing, and Valsalva manoeuvres. This results in increased blood flow and turbulence
Thus, blood from the palisade zone alternately in the intrinsic venous trunks. The dilated deep
drains into the gastric zone (i.e., into the portal sys- intrinsic veins displace the superficial venous
tem) and into the perforating zone, which extends plexus and assume a subepithelial position (Figure

Figure 6a+b. a) Oesophagocardiac junction in portal hypertension:

the submucosal veins substitute the muscularis mucosa
above the oesophagocardiac junction (thick arrow!):
sub- and intraepithelial channels or wall ectasias are
marked by triangles and represent the red colour signs
or teleangiectasias or cherry red spots or minivarices:
perforating veins (thin arrows!) E=epithelia, Lp=lamina
propria, Mm=muscularis mucosae, Subm=submu-
cosa, M.p. muscularis propria, S=serosa with perioe-
sophageal varices. b) Oesophageal varices in portal
hypertension. CRE=cherry red sports; RWM=red wale
markings; HCS=haemocystic sports (from Paquet KJ
and Kuhn R 1997 [40])

921
Clinical Research

6 a+b). These are tortuous, large variceal trunks

seen at endoscopy [37,38]. Dilatation of the intra-
epithelial veins result in the endoscopically recog-
nisable teleangiectasia, cherry red spots, red colour
signs, hemocystic sports, red wale markings or
varices on varices [37–40]. The dilatation and
increased pressure in the perforating zone influ-
ences blood to flow caudally into the palisade
zone. The palisade zone, however, must also acco-
modate the increased gastrosplenic flow associated
with portal hypertension. Consequently, blood
flow in the palisade zone is more turbulent in
patients with portal hypertension. High pressures Figure 7. Technique of paravariceal endoscopic injection for scle-
and turbulent flow in superficially located dilated rotherapy
vessels in the distal 5 cm of the oesophagus (i.e.,
the palisade and perforating zones) contribute to
the likelihood of rupture of the varices.

The current practise of endoscopic therapy for

variceal bleeding is directed at obliterating the
varices at the distal oesophagus and the oesopha-
gogastric junction which are most likely to bleed.
However, if the varices are completely obliterated,
relatively quickly new varices may develop and
cause hemorrhage from other sites. Therefore the
aim of our group is to protect the varices by mainl-
ly paravariceal injection (Figure 7) by inducing
fibrosis and scar tissue (Figure 8), thus preserving
the collateral circulation and thus preventing, par-
tially or totally the new formation of collaterals or
varices [39–43]. The eradication of the varices is
accomplished by injecting the sclerosant into or
into and around the varices to cause thrombosis of
these channels and mucosal fibrosis around them,
or by ligation of these variceal channels with endo-
scopically applied rubber bands. However, as
stressed above, the incompetent perforating veins
which connect the intrinsic venous trunks to the
perioesophageal veins are not effectively oblitera-
ted and new collateral channels my from in the
distal oesophagus.
Figure 8. Histologic specimen of a segment of the distal oesophagus
PATHOPHYSIOLOGY OF VARICEAL following paravariceal endoscopic sclerotherapy: the majori-
HAEMORRHAGE ty of the submucosal veins are preserved and still perfused

The earlier concept of erosion theory, which sug-

gested that the oesophageal varices bleed due to cular tissue, which gives support to the vessel.
external trauma to thin-walled varices, has been Tissue support is particularly important in the
abandoned because of lack of supporting objective oesophageal and rectal varices since surrounding
evidence. Factors involved in the variceal hemor- tissue provides very little support to the variceal
rhage appear to be multiple. A certain increase in wall. With continuing increase in intravariceal pres-
the portal venous pressure is required for the initia- sure, the varices expand in size and the variceal
tion of oesophageal collateral formation. The vol- wall thickness decreases. Bleeding probably occurs
ume of blood traversing the collateral vein has a when the expanding force can no longer be
role in the enlarging collaterals as does the perivas- counter-balanced by the variceal wall tension; at

922
 Paquet KJ – Causes and pathomechanisms of oesophageal varices…

pressure, which results in the opening of collater-

als. With progression of portal hypertension , the
variceal blood flow and intravariceal pressure
increases. The variceal size increases and the wall
becomes thinner. Furthermore increase in portal
pressure, or a defect in the variceal wall leads to
variceal hemorrhage. Portal pressure and blood
flow are not static and vary markedly with various
physiological stimuli. Portal flow increases tran-
siently after meals because of postprandial hyper-
aemia. Similarly, fluctuations in portal pressure has
been observed during circadian rhythm: portal
pressure being higher at night and lower during
afternoon and evening (Garcia-Pagan et al 1994)
[45]. These variations in portal pressure may influ-
Figure 9. Relative position and size of varices with respect to the ence the onset of bleeding in patients with already
oesophagus and oesophageal lumen. At equal transmural elevated baseline portal pressure.
pressure (TP), the tension (T) developed will be greater in T1
than in T2 due to increased radius (r) and smaller wall thick- NATURAL HISTORY OF VARICEAL BLEEDING
ness (w).
Oesophageal varices

this point the varices rupture and bleed. (Polio and The finding of oesophageal varices in cirrhotic
Groszman 1986) [44]. The factors regulating the patients carries a grim prognosis [46]. It is estimat-
tension experted by the wall of varices to contain ed that 50% of these patients will bleed from their
the expanding force can be described according to varices at some time in their lifes [47,48]. Factors
the LAPLACE-LAW (Figure 9). that can predict which patient will bleed and when
this event will occur have been partially elucidated.
T = TP x r/w The magnitude of the elevation of portal pressure
appears predictive in that patients, whose portal
in which T is the wall tension, TP the transmural pressure is above 10–12 mmHG, have an
pressure, r the radius and w the wall thickness of increased risk of bleeding compared to those portal
the varices. The transmural pressure is the differ- pressures are below this level [27,49]. This has
ence between intravariceal pressure and the pres- been shown by our group in two controlled ran-
sure in the oesophageal lumen. The entire varix is domised trials [38,52] and by others. However, an
in equilibrium between the outwardly directed elevation in portal pressure does not necessarily
flow (TP x r/w) and the inwardly directed force of mean that the pressure in oesophageal varices is
the wall tension. The inwardly directed force is also high, because individual patients may differ in
augmented by the surrounding tissue structures, the type of collateral circulation [50]. As has been
which provide reinforcement for the vessel wall. demonstrated by our group, too and later by
japanese investigators, teleangiectasia, varices on
The larger variceal size multiplies the deleterious varices, red colour signs (i.e., cherry red spots, red
effect of high intravariceal pressure by increasing wale markings, or hemocystic spots) and blue
the expanding force. When varix distension varices, which may indicate thin overlying mucosa
increases, the radius increases and wall thickness have been associated with a history of variceal
decreases (Figure 9). The wall tension increases in bleeding [38,40]. This relation is still debatable. In
an attempt to contain the expanding force. Since a prospective trial, only 19% of patients with these
venous structures have little musculature, this limits variceal stigmata bled from oesophageal varices
their capacity to develop wall tension. Eventually, over a 2 year period [51]. However, our group was
the expanding force in an enlarging varix exceeds able to identify a higher frequency of bleeding up
the limit of wall tension and variceal rupture to 80% over a period of three years by combining
occurs. the large varices with endoscopic signs of bleeding
and a wedged hepatic pressure gradient over 16
The consequence of events leading to portal mm HG [52]. Thus large varices with endoscopic
variceal bleeding are initiated by increased portal signs are associated with an increased likelihood of

923
Clinical Research

cirrhosis has been marked by our group and

recently confirmed by others.

Histological, radiological and Doppler-flow studies

have demonstrated that the red colour signs
detected during endoscopy correlate with blood
flow, both through dilated subepithelial communi-
cating veins, which connect the deeper extrinsic
veins of the oesophagus to the more superficial
submucosal veins [31–36]. Dilated subepithelial
veins which appear as raised red areas (cherry red
spots, teleangiectasias, minivarices, varices on
varices, red wale markings) are generally about 1 to
2 mm in diameter and, if present, lie on top of the
large dilated subepithealial or submucosal vessels,
which are usually more than 5 mm in size. Under
certain conditions, the muscularis mucosa may
atrophy in some areas, resulting in a confluence of
lamina propria and submucosa. Then, enlarged
submucosal veins can appear above the muscularis
mucosae. The hemocystic spot, which is approxi-
mately 4 mm in size, appears to represent blood
flow coming from deeper extrinsic veins of the
oesophagus straight out toward the lumen through
communicating veins into the more superficial sub-
mucosal veins. Doppler flow studies have also
demonstrated the possibility of causing stasis at the
side of communicating veins, eruption and massive
hemorrhage of these vulcano-like lesions. When
red colour signs are present, there is higher fre-
quency of various lesions (cherry red spots, mini-
varices or teleangiectasias can be seen and red
wale markings in 90%, and hemocystic sports in
80%). These endoscopic signs may represent chan-
nels in the epithelium rather than in the subepithe-
lium.

When bleeding from oesophageal varices occurs, it

usually stops spontaneously, at least temporarily, in
up to two thirds of patients. Still, 30% to 40% of
these patients are at risk for rebleeding within 2 to
Figure 10. Endoscopic grading (classification) of oesophageal varices 3 days and 60% rebleed within one week.
according to our group (From Paquet, 1982 [38]) Mortality in the first week following the index
bleed is around 25%, with rebleeding and liver fail-
ure accounting for most of the deaths [40].
variceal hemorrhage. This has been described in
our classification of oesophageal varices (Figure 10) Most studies have found a correlation between the
included in our first controlled randomised trial extent of liver failure in patients with bleeding
published in 1982 [38]. An average of 30%, and as oesophageal varices, as expressed by the Child-
many as 83% of patients who have large varices Pugh-Classification, and mortality [53,54]. Survival
bleed from them in the 25 months following their statistics are best for patients with minimal liver
discovery. dysfunction (Child’s group A) and worst for the
sickest patients (Child’s group C). For the Child’s
Variceal size as a prognostic indicator of bleeding group-C-patients, the one month mortality is
and survival in patients with mainly alcoholic liver greater than 45%, 1-year survival is 35%, and 2-

924
 Paquet KJ – Causes and pathomechanisms of oesophageal varices…

year survival is 23%. Early rebleeding may be an Benhamou [56]. These varices may be less likely to
overt reflection of the degree of liver failure and rupture because they are found deeper in the gut
has been correlated with death within 30 days. wall than oesophageal varices and the frequency is
counted not more than 10%. Duodenal varices are
Graham and Smith [46] have shown that bleeding- often found in individuals with extrahepatic portal
associated mortality rate is highest in the first days hypertension where the varices serve to bypass the
to weeks after the variceal bleed and returns to obstructed segment. In patients with intrahepatic
baseline by 3 to 5 months. Therefore, therapeutic portal hypertension, duodenal varices form
interventions are likely to have the most influence between the afferent branches of the portal vein,
on survival if introduced in the early period follow- such as the superior mesenteric vein, and the infe-
ing the bleed. rior vena cava [57,58].

Gastric varices Patients with bleeding duodenal varices commonly

present with haematemesis or melaena. During
In patients with portal hypertension, gastric varices emergency endoscopy, these varices may be diffi-
are predominantly supplied by dilated short gastric cult to identify and concomitant oesophagogastric
veins and are frequently found with oesophageal varices that are more easily visualised may be
varices. Gastric varices may form as a result of assumed to be the source of bleeding. Angiography
increased resistance to cephalad flow through the or dynamic CT scan may be needed to confirm the
palisade zone at the gastrooesophageal junction in diagnosis [57–58].
these patients. Isolated gastric varices are found in
patients with splenic vein thrombosis. Portal pressure

Gastric varices may be more difficult to identify at Without doubt, there is a pathogenic correlation
endoscopy because there are generally situated between elevated portal pressure and variceal
deeper than oesophageal varices and may resem- hemorrhage. However, the exact correlation is still
ble rugal folds. In actually bleeding patients, the a matter of controversy.
fundus is frequently obscured by a pool of blood
and small gastric varices may not be easily detect- First of all, certain levels of portal pressure or por-
ed. In these cases the patients should be put in a tohepatic gradient of 10/12 mm HG are required
right site position to be more accurate and sure in for the development of oesophageal varices
detecting the varices and source of bleeding. [59,60]. At levels lower than these, the presence of
varices is unlikely. In patients with liver disease in
In much of the literature, gastric varices are report- an early stage, the presence of varices correlates to
ed together with oesophageal varices rather than as the degree of portal hypertension. However, in
a separate entity. Therefore, the incidence, fre- patients with established portal hypertension the
quency of bleeding, and natural history of gastric relation between variceal hemorrhage and the
varices have not been defined precisely. Incidence level of portal venous pressure is still questioned
rate of 16% to 70% have been reported in patients [61,62]. For this controversy the following reasons
with portal hypertension and cirrhosis [29]. are responsible:
Whether mere presence of gastric varices influ-
ences survival is unknown. However, patients who 1. There are still methodological problems in mea-
bleed from gastric varices present a therapeutic suring portal pressure, even when the method of
challenge because they are part of an extensive occluded wedged hepatic venous pressure is
collateral network, making hemostasis and ultimate used.
obliteration difficult [55].
2. The pressure in the oesophageal varices may be
Duodenal varices more important for the risk of bleeding than the
portal pressure because formation of sponta-
Apart from oesophagogastric varices, duodenal neous shunts can be followed by pressure
varices are one of the more commonly reported reduction, equalisation, compensation or pres-
digestive tract varices in portal hypertension. The sure balance [63].
incidence of these varices is not known, but the
duodenal varices represented one third of the 3. The oesophageal varices are localised quite dis-
bleeding ectopic varices reported by Lebrec and tant from the portal vein, so that the possibility

925
Clinical Research

Table 2. Heinz Kalk-Classification of hepatocellular function - a modi- the hypothesis of explosion as a course of variceal
fied Child-Pugh-Classification (proposed by our group) rupture and abrupt variceal hemorrhage.

Observation Points Points Points Last but not least, other predisposing factors for
variceal hemorrhage like ascites, coagulation disor-
Nutritional status exellent good poor ders, local elevated fibrogenolysis, the last not to
Ascites none moderate marked common in prehepatic block without liver disease,
easily controlled poorly controlled are all included in the Child-Pugh-classification and
Neurological disorder none I-II III-IV their modification by our group (Table 2), responsi-
(encephalopathy) ble as well for the risk of bleeding, liver failure and
Serum bilirubin (mg/100ml) < 2.0 2.0 – 3.0 > 3.0 thus prognosis [48,53,54].
Serum albumin > 3.5 3.5 – 3.5 < 3.0
Prothrombin concentr. > 75% 50% – 75% < 50% Therefore a good clinician or hepatologist should
(Quick - test) be aware of all these variables when he is responsi-
child A = 6-8 point; child B = 9-11 point; child C = 12 or more point ble for a patient with portal hypertension.

REFERENCES:
of vascular resistance in this area can not be com-
pletely excluded. 1. Greenway CV, Stark RD: Hepatic vascular bed. Physiological Review,
1971; 51: 23-65
Rigau et al [59] found that portal pressure in 2. Genecin P, Groszman RJ: Portal Hypertension. In: Schiff E, Schiff L
patients with oesophageal variceal bleeding was (eds): Diseases of the Liver. JB Lippincott Company, Philadelphia
significant higher than in those patients without 1993, pp 935-973
bleeding. Staritz et al [63] compared 14 patients 3. Popper H, Zak FG: Pathological aspects of cirrhosis. Am J Med,
with compensated liver cirrhosis and status post- 1958; 24: 593-625
variceal hemorrhage with 15 similar patients with- 4. Baldus WP, Hoffbauer FW: Vascular changes in the cirrhotic liver as
out bleeding and determined that patients with studied by injection technique. Am J Digest Dis, 1993; 8: 689-692
bleeding had a median intravariceal pressure of 22 5. Bhathal PS, Groszman HJ: Reduction of increased portal vascular
mmHG and patients without bleeding of 15 resistance of the isolated perfused cirrhotic rat liver by vasodilators. J
mmHG. Hepatol, 1985; 1: 325-337
6. Beker S, Valencia-Parparcen J: Portal hypertension syndrome. A com-
Thus, oesophageal pressure of at least 20 mmHG parative analysis of bilharzial fibrosis and hepatic cirrhosis. Am J
represents a high danger of bleeding. Furthermore, Digest Dis, 1968; 13: 1047-1054
a positive correlation between pressure in 7. Van Leeuwen DJ, Howe SC, Scheuer PJ et al: Portal hypertension in
oesophageal varices and their size was found [64]. chronic hepatitis: relationship of morphological changes. Gut, 1991;
In patients with alcoholic cirrhosis this Reynolds- 31: 2339-343
Rector-group measured the portal pressure by 8. Schaffner F, Popper H: Capillarization of the hepatic sinusoids in
means of an indwelling hepatic vein balloon man. Gastroenterology, 1963; 76: 704-709
catheter during and after the bleeding episodes. 9. Reynolds TB, Hidmura R, Mitchell H et al: Portal hypertension with-
Portal pressure was significantly higher than 16 out cirrhosis in alcoholic liver disease. Ann Int Med, 1969; 70: 497-
mmHG in patients who continued to bleed or 506
rebleed compared to patients remaining stable. 10. Lieber CS, Zimmon DS, Kessler RE et al: Portal hypertension in
They stated that continuous monitoring of portal experimental alcoholic liver injury. Clin Res, 1976; 24: 478 (a)
pressure in patients with bleeding oesophageal 11. Brion E, Brenard R, Pariente EA et al: Sinusoidal portal hypertension
varices due to alcoholic cirrhosis is safe and feasi- in hepatic amyloidosis. Gut, 1991; 32: 227-230
ble and permits rapid stratification of the risk of 12. Pinzani M, Failli P, Ruocco C et al: Fat-storing cells as liver-specific
bleeding, continuous bleeding and rebleeding. pericytes-spatial dynamics of agonist-stimulated intracellular calcium
Nearly all authors involved in measuring transients. J Clin Invest, 1992; 90: 642-646
oesophageal variceal pressure by different methods 13. Mc Clure RR, Woodward M: Contractile fibroblasts in chronic alco-
agree in the statement that oesophageal pressure is holic cirrhosis. Gastroenterology, 1979; 76: 704-709
more elevated in large varices than in small one 14. Rubanyi GM: Endothelium-Derived relaxing and contracting factors.
[65–67]. Furthermore, this pressure is influenced J Cell Biochem, 1990; 46: 27-36
by respiration, the Valsalva manoeuvre, vomiting 15. Vane JR, Anggard EE, Botting RM: Regulatory functions of vascular
and pressing the abdomen to a great extent, as endothelium. N Engl J Med, 1990; 323: 27-36
mentioned above [65–67]. All these data underline

926
 Paquet KJ – Causes and pathomechanisms of oesophageal varices…

16. Mittal MK, Gupta TK, Lee FY et al: Nitric oxide modulates vascular 37. Beppu K, Inokuchi K, Koyanagi N et al.: Prediction of variceal haem-
tone in normal rat liver. Am J Physiol, 1994; 267: G416-G422 orrhage by oesophageal endoscopy. Gastrointestinal Endoscopy, 1981;
17. Gupta TK, Chung M, Sessa WC et al: Impaired endothelial function 27: 213
in the intrahepatic circulation in cirrhosis. Hepatology, 1995; 22(4): 38. Paquet KJ: Prophylactic endoscopic sclerosing treatment of the
156A oesophageal wall in varices – a prospective controlled randomized
18. Gupta TK, Groszman RJ: Administration of L-arginine, the physio- trial. Endoscopy, 1982; 14: 4
logical precursor of nitric oxide, reduces portal perfusion pressure and 39. Paquet KJ, Fleig W: Sclerotherapy of oesophageal varices. In: Papp
ameliorates hepatic vascular hyperreactivity in experimental cirrhosis. UP (ed): Endoscopic Control of Gastrointestinal Haemorrhage. CRC
Hepatology, 1994; 20(4): 200A Press, Boca Raton 1981, Florida, USA
19. Rockey DC, Weisiger RA: Endothelin induced contractility of stellate 40. Paquet KJ, Kuhn R: Prophylactic endoscopic sclerotherapy in patients
cell from normal and cirrhotic rat liver: implication of regulation of with liver cirrhosis and oesophageal varices. Hepato-Gastroenterology,
portal pressure an resistance. Hepatology, 1996; 24(1): 233-240 1997; 44: 625-636
20. Columbato LA, Albillos A, Groszman RJ: Temporal relationship of 41. Paquet KJ, Raschke E: Experiences in the management of the acute or
peripheral vasodilatation, plasma volume expansion and the hyperdy- impending haemorrhage from oesophageal varices using the oesophago-
namic circulatory state in portal hypertensive rats. Hepatology, 1992; scopic sclerosing method. In: International Symposium on acute gas-
15: 323-328 trointestinal haemorrhage. Prague-Carlsbad, Karger, Basel 1972, pp 184
21. Liebermann FL, Reynolds TB: Plasma volume in cirrhosis of the liver. 42. Paquet KJ, Raschke E: Management of haemorrhage from
J Clin Invest, 1967; 46: 1279-1308 oesophageal varices using the endoscopic sclerosing method. Ann
22. Halvorsen JF, Myking AO: The portosystemic collateral pattern in the Surg, 1973; 117: 99
rat. Eur Surg Res, 1974; 6: 183-195 43. Paquet KJ, Oberhammer E: Sclerotherapy of bleeding oesophageal
23. Sikuler E, Groszman RJ: Haemodynamic studies in long- and short- varices by means of endoscopy. Endoscopy, 1978; 10: 7-12
term portal-hypertensive rats; the relation of systemic glucagon levels. 44. Polio J, Groszman RJ: Haemodynamic factors involved in the devel-
Hepatology, 1996; 6(3): 414-416 opment and rupture of oesophageal varices: a pathophysiological
24. Lee FY, Colombato LA, Albillos A et al: Administration of N-omega- approach to treatment. Sem Liver Dis, 1986; 6: 318-331
nitro-L-arginine ameliorates porto-systemic shunting in portal-hyper- 45. Garcia-Pagan JC, Feu F, Castells A et al.: circadian variations of por-
tensive rats. Gastroenterology, 1993; 105: 1464-1470 tal pressure and variceal haemorrhage in patients with cirrhosis.
25. Lee FY, Colombato La, Albillos A et al. (1993) N-nitro-L-arginine Hepatology, 1994; 5: 595-601
administration corrects vasodilatation and systemic capillary hypoten- 46. Graham DY, Smith JL: The course of patients after variceal haemor-
sion, and ameliorates plasma volume expansion and sodium retention rhage. Gastroenterology, 1981; 80: 800
in portal hypertensive rats. Hepatology, 1993; 17: 84-90 47. Koff RS: Benefit of endoscopy in upper gastrointestinal bleeding in
26. Mosca P, Lee FY, Kaumann AJ et al.: Pharmacology of portal-sys- patients with liver disease. Dig Dis Sci, 1981; 26: 12
temic collaterals in portal hypertensive rats: role of endothelium. Am J 48. Galambos JT: Oesophageal variceal haemorrhage: diagnosis and an
Physiol, 1992; 262: G544-G550 overview of treatment. Semin Liver Dis, 1982; 2: 211
27. Garcia-Tsao G, Groszman RJ, Fisher RL et al.: Portal pressure, pres- 49. Lebrec D, De Fleury, Rueff B et al.: Poral hypertension, size of
ence of gastrooesophageal varices and variceal bleeding. Hepatology, oesophageal varices, and risk of gastrointestinal bleeding in alcoholic
1985; 5: 419 cirrhosis. Gastroenterology, 1980; 79: 1139
28. Butler H: Gastro-oesophageal haemorrhage in hepatic cirrhosis. 50. Rector WG, Reynolds TB: Risk factors for haemorrhage from
Thorax, 1953; 7: 159 oesophageal varices and acute gastric erosions. Clin Gastroenterol,
29. Sarin SK, Schadev G, Nanda R et al: Endoscopic sclerotherapy in the 1985; 14: 139
treatment of gastric varices. Br J Surg, 1998; 75: 747 51. Inokuchi K: Prophylactic portal non-decompressive surgery in patients
30. Butler H: Veins of oesophagus. Thorax, 1951; 6: 276 with oesophageal varices. Ann Surg, 1984; 200: 61
31. Kitano S, Terblanche J, Kahn J et al.: Venous anatomy of the lower 52. Paquet KJ, Kalk JF, Klein CP et al.: Prophylactic sclerotherapy for
oesophagus in portal hypertension: practical implications. Br J Surg, oesophageal varices in high-risk cirrhotic patients selected by endo-
1986; 73: 525 scopic and hemodynamic criteria. Endoscopy, 1994; 26: 734-740
32. Vianna A, Hayes PC, Moscoso G et al. Normal venous circulation of 53. De Franchis R, Primignani M: Why do varices bleed? Gastroenterol
the gastrooesophageal junction. Gastroenterology, 1987; 95: 876 Clin North Am, 1992; 21: 85
33. Spence RAJ, Terblanche J: Venous anatomy of the lower oesophagus: 54. Burroughs A, Mezzanotte G, Philipps A et al.: Cirrhotics with varceal
a new perspective on varicose. Br J Surg, 1987; 74: 659 haemorrhage: the importance of the time interval between admission
34. Hashizume M, Kitano S, Sugimachi K et al: Three-dimensional view and the start of analysis for survival and rebleeding.
of the vascular structure of the lower oesophagus in clinical portal 55. Trudeau W, Prindiville T: Endoscopic injection sclerosis in bleeding
hypertension. Hepatology, 1988; 8: 1482 gastric varices. Gastrointest Endosc, 1986; 32: 264
35. Mc Cormack TT, Rose JD, Smith PM et al: Perforation veins and 56. Lebrec D, Benhamou JP (1985) Ectopic varices in portal hyperten-
blood flow in oesophageal varices. Lancet, 1983; 24: 1442 sion. Clin Gastroenterol, 1985; 14: 105
36. Noda T: Angioarchitectural study of oesophageal varices – with spe- 57. Sauerbruch T, Weinzierl M, Dietrich HP et al.: Sclerotherapy of a
cial reference to variceal rupture. Virchows Archiv, 1984; 404: 381 bleeding duodenal varix. Endoscopy, 1982; 14: 187

927
Clinical Research

58. Paquet KJ, Lazar A, Bickart J: Massive and recurrent gastrointestinal 63. Staritz M, Meyer zum Büschenfelde KH: The endoscopic measure-
haemorrhage due to jejunal varices in afferent loop – diagnosis and ment of intravascular pressure and flow in oesophageal varices. J
management. Hepato-Gastroenterology, 1994; 41: 276-277 Hepatology, 1988; 7: 126-131
59. Rigau J, Bosch J, Bordaz JM et al.: Endoscopic measurement of 64. Bosch J, Bordas JM, Rigau J et al.: Non invasive measurement of the
variceal pressure cirrhosis. Correlation with portal pressure and pressure in oesophageal varices using and endoscopic gauge: compari-
variceal haemorrhage. Gastroenterology, 1989; 96: 873-880 son with the measurements by variceal puncture in patients undergo-
60. Vialett A, Magelau D et al.: Haemodynamic evaluation of patients ing endoscopic sclerotherapy. Hepatology, 1986; 6: 667-672
with intrahepatic portal hypertension. Relationship between bleeding 65. Gersch, P: Endoscopic non-invasive manometry of oesophageal
varices and the portohepatic gradient. Gastroenterology, 1975; 68: varices: prognostic significance. Am J Surg, 1982; 144: 528-530
1297-1300 66. Gross E, Erhard J: Endoskopisch geführte Druckmessung in distalen
61. Koch H, Henning H, Grimm H et al.: Prophylactic endoscopic scle- Ösophagusvarizen. Dtsch Med Wschr, 1987; 112: 125-127
rotherapy of oesophageal varices – results of a prospective controlled 67. Ready JB, Robertson AD, Gow JS et al.: Assessment of the risk of
study. Endoscopy, 1986; 18: 40-43 bleeding from oesophageal varices by continous monitoring of portal
62. Sauerbruch T, Kleber G, Gerdes A et al.: Prophylaxis of first variceal pressure. Gastroenterology, 1991; 100: 1403-1410
haemorrhage in patients with liver cirrhosis. Klein Wschr, 1986; 64:
1267-1275

928
Index Copernicus
Global Scientific Information Systems
for Scientists by Scientists

www. IndexCopernicus.com

EVALUATION & BENCHMARKING

PROFILED INFORMATION
NETWORKING & COOPERATION
VIRTUAL RESEARCH GROUPS
GRANTS
PATENTS
CLINICAL TRIALS
JOBS
STRATEGIC & FINANCIAL DECISIONS

Index IC Scientists IC Virtual Research Groups [VRG]

Effective search tool for Web-based complete research
Copernicus collaborators worldwide.
Provides easy global
environment which enables researchers
to work on one project from distant
networking for scientists. locations. VRG provides:
integrates C.V.'s and dossiers on selected
scientists available. Increase
customizable
 and individually
self-tailored electronic research
your professional visibility.
protocols and data capture tools,
statistical
 analysis and report
IC Journal Master List IC Patents creation tools,
Scientific literature database, Provides information on patent profiled information on literature,
including abstracts, full text, registration process, patent offices 
publications, grants and patents
and journal ranking. and other legal issues. Provides related to the research project,
Instructions for authors links to companies that may want
available from selected journals. to license or purchase a patent. administration
 tools.

IC Grant Awareness
IC Lab & Clinical Trial Register
Need grant assistance?
IC Conferences
Step-by-step information on Provides list of on-going laboratory
Effective search tool for how to apply for a grant. Provides or clinical trials, including
worldwide medical conferences a list of grant institutions and research summaries and calls for
and local meetings. their requirements. co-investigators.