REVIEW

Achintya D. Singh, MD, MBBS Simon R. Mucha, MD Christina C. Lindenmeyer, MD

Department of Gastroenterology and Hepatology, Department of Critical Care Medicine, Respiratory Co-director, Cleveland Clinic Medical Intensive Liver Unit,
MetroHealth Medical Center, Case Western Institute, Cleveland Clinic; Clinical Instructor, Department of Gastroenterology, Hepatology, and Nutrition,
Reserve University, Cleveland, OH Cleveland Clinic Lerner College of Medicine of Digestive Disease and Surgery Institute, Cleveland Clinic,
Case Western Reserve University, Cleveland, OH Cleveland, OH; Associate Director of Basic Science Education,
and Assistant Professor, Cleveland Clinic Lerner College of
Medicine of Case Western Reserve University, Cleveland, OH

Cirrhotic coagulopathy:
A rebalanced hemostasis
ABSTRACT oagulopathy—characterized by pro-
Cirrhosis has been regarded as a hypocoagulable state
C longed prothrombin time, elevated inter-
national normalized ratio (INR) of the prothrom-
associated with an increased risk of bleeding. But bin time, prolonged activated partial thromboplas-
patients with cirrhosis also have a high incidence of tin time, low fibrinogen levels, and low platelet
thrombotic complications, challenging this dogma. We counts—is a hallmark of advanced cirrhosis. Tra-
now recognize that in cirrhosis there is a simultaneous ditionally, cirrhosis has been considered a hypoco-
decrease in both clotting and anticlotting factors, leading agulable state in which the risk of life-threatening
to a new equilibrium. Conventional coagulation tests bleeding complications is increased.1–3
such as the platelet count and prothrombin time do not Evidence of this comes from the PRO-
assess the reduced anticoagulation factors in cirrhosis LIVER study,4 which prospectively followed
280 patients with cirrhosis for a median of 1,129
and overestimate the bleeding risk, and any intervention days. Significant bleeding events occurred in
based on these test results can lead to thrombotic com- 5.45% of patients per year.4 The bleeding rate
plications. This article reviews the changes in hemostasis is higher in patients with advanced liver dis-
associated with cirrhosis, newer tests for assessing coag- ease who need to be admitted to the hospital
ulation, and preprocedural minimization of coagulopathy. because of acute decompensation of their liver
disease, or for patients with cirrhosis who need
KEY POINTS to be admitted to the intensive care unit for
The rebalanced hemostasis of cirrhosis is a delicate equilib- any reason.3,5 Most of these bleeding events
rium of antithrombotic and prothrombotic changes associ- are gastrointestinal and most are thought
ated with decreased synthetic liver function, inflammation, to be related to elevated portal pressures.4,5
and endothelial and platelet activation related to cirrhosis. Importantly, markers of coagulopathy such
as elevated INR, thrombocytopenia, and low
There is no evidence to support routine transfusion of fibrinogen levels have not been shown to
correlate with or predict the risk of bleeding
blood products to "correct" coagulopathy before low-risk events accurately.6
procedures, since this does not decrease procedure-specific However, patients with cirrhosis also have
bleeding risk and is itself associated with significant risk. a high incidence of thrombotic complications
such as portal vein thrombosis and venous
Viscoelastic tests such as thromboelastography may thromboembolism, which are independently
better reflect the true state of cirrhotic hemostasis, but associated with significant morbidity, acute
further studies are needed to establish validated trans- hepatic decompensation, and death.7,8 The
fusion thresholds. incidence of portal vein thrombosis in patients
with cirrhosis has varied widely in different
studies, owing to differences in the popula-
doi:10.3949/ccjm.89a.21018 tions studied, but it is higher than in patients

CLEVELAND CLINIC JOURNAL OF MEDICINE VOLUME 89 • NUMBER 9 SEPTEMBER 2022 523

Downloaded from www.ccjm.org on April 19, 2024. For personal use only. All other uses require permission.
CIRRHOTIC COAGULOPATHY

ors
n fact
gu latio
icoa
Ant

III
bin
t i t hrom
ÈAn C
ein
ÈProt
e i nS tor
Pro t n tiva
È i n oge gen ac
m o
È Plas min
rs plas
to T issue
fac Ç 
n
latio
a gu
co
Pro
tor
l ets 13 d fac r
t e TS ran
Pla ito
È DAM illeb inhib
A
È on W VIII tor
v r iva
Ç acto VII n act
F r e
Ç acto inog
F
È lasm
P
È
Acute kidney injury
Invasive procedures
Acute decompensation
of cirrhosis Infections

Clotting
Bleeding complications
complications

2022

Figure 1. Coagulation and anticoagulation in patients with cirrhosis are rebalanced due to simultaneous
decreases in clotting and anticlotting pathways. However, this balance is dynamic, and concomitant condi-
tions such as infection and acute kidney injury can tip the balance, resulting in a clotting or bleeding compli-
cation. (ADAMTS13 = a disintegrin and metalloproteinase with thrombospondin type 1 motifs, member 13)

without cirrhosis.9,10 A rate ranging between 3.2% lenges the notion that patients with cirrhosis are
and 4.1% at 1 year after diagnosis is often cited and “autoanticoagulated” and highlights the need for
increases over time.11,12 a more nuanced evaluation of the coagulopathy of
In a large case-control study, the relative risk of cirrhosis.
venous thromboembolism in patients with cirrhosis In fact, the liver synthesizes most proteins of the
was found to be 1.74 (95% confidence interval [CI] coagulation system. Cirrhosis results in a simultaneous
1.54–1.95) compared with patients without liver decrease in both procoagulant and anticoagulant fac-
disease.13 These findings were echoed by data from tors, resulting in a delicate state of rebalanced hemo-
the Multiple Environmental and Genetic Assess- stasis,15 metaphorically illustrated in Figure 1. This
ment study,14 which showed that in hospitalized coagulation profile is unique to the individual patient
patients, liver disease was associated with significantly and is influenced by the etiology of liver disease, dis-
increased risk of venous thromboembolism (adjusted ease severity, acute illnesses, and ongoing therapy.16,17
odds ratio [OR] 1.7, 95% CI 1.0–2.9).14 Conventional coagulation tests such as prothrombin
This increased risk of thrombotic events chal- time, INR, and platelet count measure procoagulant

524 CLEVELAND CLINIC JOURNAL OF MEDICINE VOLUME 89 • NUMBER 9 SEPTEMBER 2022

Downloaded from www.ccjm.org on April 19, 2024. For personal use only. All other uses require permission.
 SINGH AND COLLEAGUES

1. Platelet activation 2. Coagulation 3. Fibrinolysis

Plasma-kallikrein Pre-kallikrein
HK
Extrinsic pathway
HK of coagulation
XII XIIa
Platelets
Tissue factor
Intrinsic pathway
of coagulation Ca 2+
(contact system) XI XIa

VIIa VII
Ca 2+
Fibrinolysis
IX IXa
VIII VIIIa
uPA, tPA
Ca 2+
X Xa
Activated
platelets V Va
Fibrinogen Plasminogen Plasmin
Ca 2+
II IIa
(Prothrombin) (Thrombin) Fibrin

Fibrin
Cross-linked
degradation
XIII XIIIa fibrin clot
Common pathway of coagulation products
2022

Figure 2. The 3 phases of coagulation, involving a range of clotting factors (as Roman numerals).
HK = high-molecular-weight kininogen; tPA = tissue plasminogen activator; uPA = urokinase plasminogen activator

factors but not the anticoagulation factors, and there- and cross-linking of fibrin polymers.
fore they are inadequate to accurately assess bleeding • Fibrinolysis or clot dissolution.
risk and guide management.2,11 Cirrhosis affects all 3 phases, leading to a delicate
The purpose of this review is to elucidate the new equilibrium—the rebalanced hemostasis of cir-
current understanding of coagulopathy of cirrhosis, rhosis. This new balance is easily disturbed and tipped
how to assess it, and how to manage bleeding risk in toward either bleeding or thrombosis by acute events
patients about to undergo invasive procedures. such as infection, renal failure, and invasive proce-
dures with or without prophylactic transfusions.
■ PATHOPHYSIOLOGY OF HEMOSTASIS
A new balance in platelet activation
The normal hemostatic process comprises 3 phases Fewer platelets. Thrombocytopenia is common
(Figure 2): in cirrhosis and portal hypertension, likely due
• Platelet activation: When the vessel wall is in- to increased platelet destruction, reduced hepatic
jured, subendothelial collagen and tissue factor synthesis of thrombopoietin, and increased splenic
are exposed, triggering platelet activation and pri- aggregation (which further increases thrombopoietin
mary hemostasis with adhesion of the initial plate- clearance).18 Tripodi et al19 found that platelet counts
let plug through interactions with von Willebrand as low as 60 × 109/L in plasma from patients with cir-
factor, factor VIII, glycoprotein IIb/IIIa receptors, rhosis were still sufficient to yield in vitro thrombin
and fibrinogen. generation similar to that in plasma from healthy
• Coagulation: Sequential activation of prothrom- controls with normal platelet counts. This highlights
botic coagulation factors leads to thrombin acti- the importance of qualitative alterations in platelet
vation, thrombus formation, and thrombus stabi- activation beyond the quantitative decrease in plate-
lization through conversion of fibrinogen to fibrin let counts in cirrhosis.

CLEVELAND CLINIC JOURNAL OF MEDICINE VOLUME 89 • NUMBER 9 SEPTEMBER 2022 525

Downloaded from www.ccjm.org on April 19, 2024. For personal use only. All other uses require permission.
CIRRHOTIC COAGULOPATHY

More von Willebrand factor. In contrast to the factor-factor VIIa complex and facilitates degradation
coagulation factors synthesized by the liver, von of factors V and VIIIa. Tissue factor pathway inhibitor
Willebrand factor is produced, stored, and released has been shown to be reduced in cirrhosis.
by the vascular endothelium. Its levels are preserved Moreover, the key anticoagulant proteins acti-
or even increased in cirrhosis.20 Levels of von Wille- vated protein C and protein S have been similarly
brand factor have also been shown to progressively shown to be reduced in cirrhosis. Activated protein C
increase with more advanced liver disease. Lisman et with protein S as a cofactor inhibits activated factors
al20 reported that, compared with healthy controls, V and VII and thus thrombin formation. The activ-
patients with Child-Pugh class A cirrhosis had von ity of protein C is regulated in the endothelium by
Willebrand factor antigen levels 380% higher, those thrombomodulin, and “thrombomodulin resistance”
with class B cirrhosis had levels 500% higher, and has been demonstrated in plasma from patients with
those with class C cirrhosis had levels 760% higher. cirrhosis.26,27 This affirms the hypercoagulable effect
Similarly, von Willebrand factor levels were 790% of decreased hepatic synthesis of proteins C and S in
higher in patients with acute liver failure.20 vitro.
Less ADAMTS13. ADAMTS13 (a disintegrin In fact, the thrombin generation potential in
and metalloproteinase with thrombospondin type plasma from cirrhotic patients has been demonstrated
1 motifs, member 13) is a potent inhibitor of von to be similar to that of noncirrhotic patients, confirm-
Willebrand factor and is reduced in cirrhosis.21 How- ing the rebalanced state of hemostasis in cirrhosis.19
ever, data are conflicting as to the magnitude of this
decrease and its correlation with disease severity, Decreased fibrinolysis
perhaps partly because ADAMTS13 is difficult to Plasmin-mediated fibrinolysis and clot dissolution is
measure and results vary with different assays.19 the final step in hemostasis. Plasmin is activated from
Thus, the data suggest that in patients with cirrho- plasminogen by fibrin, as well as by tissue plasminogen
sis, thrombocytopenia is countered by a simultaneous activator, urokinase plasminogen activator, and acti-
increase in von Willebrand factor and a decrease in vated factor XII. Conversion of plasminogen to plas-
ADAMTS13, resulting in only mildly decreased or min is inhibited by thrombin activatable fibrinolysis
even increased platelet activity. Indeed, the excess inhibitor and plasminogen activator inhibitor.15,28 A
risk of thrombosis observed in patients with biliary decrease in the plasmin activation pathway results in
cirrhosis has been attributed to platelet activation.22 a hypofibrinolytic state, while an increase results in a
hyperfibrinolytic state.
Rebalanced coagulation vs anticoagulation Cirrhosis has been shown to be associated with
The coagulation cascade is driven by procoagulant both quantitative and qualitative changes in the
factors and inhibited by anticoagulant factors.23,24 As fibrinolytic pathway. Plasminogen levels are reduced
both types of factors are predominantly produced in in patients with cirrhosis, likely due to the combined
the liver, this phase is rebalanced in cirrhosis. effects of decreased production and increased con-
Procoagulation factors are decreased. Release of sumption related to the frequent activation of the
tissue factor from the endothelium activates factor VII, coagulation cascade from ongoing inflammation.24
forming the tissue factor-VIIa complex, which leads Oxidative stress, leading to modifications of fibrin,
to activation of factors V, IX, and X and ultimately and increased sialic acid content and altered cal-
to conversion of prothrombin to thrombin. As all cium binding lead to decreased clot permeability and
coagulation factors except for factor VIII (produced in impaired fibrinolysis.29,30 Together these qualitative
hepatic sinusoidal endothelial cells) are synthesized in and quantitative changes result in a net decrease in
hepatocytes, conventional coagulation tests that assess fibrinolysis in cirrhosis.
procoagulant factors tend to suggest a hypocoagulable
state in conditions of hepatic synthetic dysfunction Inflammation and infection can tip the balance
such as cirrhosis or liver failure: eg, the prothrombin Systemic inflammation is an important factor in
time will be prolonged and the INR will be high. the development and progression of chronic liver
Anticoagulation factors are also decreased. Anti- disease, acute liver failure, and acute-on-chronic
coagulation factors, in contrast, exert their effect in liver failure. Moreover, patients with liver disease
the endothelium and are difficult to quantify in vitro.25 are at increased risk of both primary and second-
For example, tissue factor pathway inhibitor forms a ary infections, which in turn contribute to disease
complex with activated factor X. It inhibits the tissue progression.31–33
526 CLEVELAND CLINIC JOURNAL OF MEDICINE VOLUME 89 • NUMBER 9 SEPTEMBER 2022

Downloaded from www.ccjm.org on April 19, 2024. For personal use only. All other uses require permission.
 SINGH AND COLLEAGUES

Coagulation Fibrinolysis

Alpha LY
angle

MA
R K

Time --------------> 2022

PT D-dimer
aPTT
ACT

Parameters Description Normal range Interpretation

R time Latent time from test initiation to initial clot 15–23 minutes Prolongation shows deficiency or inhibition of
formation and is approximately equal to the rate (whole blood) clotting factors (anticoagulants or autoimmune
of thrombin formation 5–10 minutes inhibitors)
(kaolin added)

Alpha angle Measures rate of clot progression 22–38 degrees Narrow or reduced angle suggests
Determined by fibrin sheath buildup (whole blood) deficiency of fibrinogen or its decreased
and cross-linking 53–67 degrees activity
(kaolin added) Low platelets counts or dysfunction in
conditions like uremia, cirrhosis can also
decrease the angle

K time Time required to achieve a particular clot 5–10 minutes Prolonged in fibrinogen deficiency or decreased
strength, ie, 20-mm amplitude (whole blood) function
1–3 minutes
(kaolin added)

Maximum Ultimate strength of the platelet clot 47–88 mm Low indicates low platelet number or function
amplitude (MA) (whole blood) or decreased fibrinogen levels or function; can
59–68 mm also be affected by other factors such as factor
(kaolin added) XIII deficiency

Coagulation index Composite indicator of the coagulation profile –3 to 3 Increased in hypercoagulable states, decreased
in hypocoagulable states

Clot lysis at 30 Percentage decrease in amplitude 30 minutes 0–7.5% Decreased in hypercoagulable states, increased
minutes (LY) after MA; indicates fibrinolysis in hyperfibrinolytic states

Figure 3. Thromboelastography is a promising test of coagulation. The horizontal axis represents time, the
vertical axis represents deflection of the thromboelastography probe. The R time is also assessed by tests
such as the prothrombin time (PT), the activated partial thromboplastin time (aPTT), and the activated
clotting time (ACT). D-dimer is used to assess fibrinolysis.
Adapted from Singh AD, Shalimar. Use of blood products and drugs before procedures in patients with cirrhosis. Clin Liver Dis (Hoboken) 2020;
16(4):153–157. doi:10.1002/cld.906, reference 36.

CLEVELAND CLINIC JOURNAL OF MEDICINE VOLUME 89 • NUMBER 9 SEPTEMBER 2022 527

Downloaded from www.ccjm.org on April 19, 2024. For personal use only. All other uses require permission.
CIRRHOTIC COAGULOPATHY

Several mechanisms link inflammation and coagu- fibrinolysis, making the absolute values difficult to
lopathy. Inflammatory cytokines lead to direct activa- interpret in the context of the rebalanced state of
tion of platelets and the endothelium, and endothelial hemostasis.
activation in turn prompts the release of tissue factor
and von Willebrand factor. Tissue factor activates the Viscoelastic tests are promising but need more study
extrinsic coagulation cascade, and increasing levels of Viscoelastic tests such as thromboelastography and
von Willebrand factor further promote platelet acti- rotational thromboelastometry provide a holistic
vation and adhesion. Increased levels of fibrinogen, evaluation of the coagulation process, assessing clot
an acute-phase reactant, can tip the balance toward formation, clot propagation, maximum clot strength,
a more hypercoagulable state. Similarly, inflamma- and fibrinolysis as a reflection of shear stress in vitro.
tion-induced expression of plasminogen activator Viscoelastic tests are performed using whole blood,
inhibitor 1 further inhibits fibrinolysis. Eventually, assessing coagulation in a more global, functional,
prolonged activation of a systemic inflammatory and potentially clinically relevant fashion than indi-
response can result in exhaustion of thrombotic and vidual coagulation parameters.
thrombolytic systems, leading to a state of consump- Thromboelastography is performed with a torsion
tive coagulopathy.34 pin suspended in an oscillating cup containing whole
These mechanisms highlight the complexity of blood.35 As the blood begins to clot, the initial platelet
coagulopathy of advanced liver disease and empha- clot and fibrin strands move this pin. The deflection
size the importance of individualized assessment and of the pin is proportional to clot strength and is dis-
management of coagulopathy in patients with cir- played graphically (Figure 3).36 The pattern is altered
rhosis and liver failure, particularly in patients with in patients with abnormal hemostasis (Figure 4).
systemic inflammation or sepsis, or both. Rotational thromboelastometry uses a stationary
cup, a rotating pin, and optical methods to measure
■ ASSESSING CLOTTING AND ANTICLOTTING clot formation instead of the shearing forces used by
IN CIRRHOSIS thromboelastography. It is considered less vulnerable
to movement and vibration.37,38
An accurate assessment of the coagulation system is Viscoelastic tests were developed as point-of-care
paramount in the clinical management of cirrhosis. tests to provide rapid results during surgery or in the
An ideal test should evaluate both the clotting and trauma bay—in 15 to 20 minutes, compared with
the anticlotting pathways to provide an accurate conventional tests, which require significantly longer
assessment of hemostasis to guide therapy. turnaround times (eg, 45–60 minutes for prothrombin
Conventional tests assess only clotting time and INR).
and may overestimate bleeding risk Viscoelastic testing is well established in trauma
The conventional tests for assessing coagulation care and cardiothoracic surgery, where its use has
are the prothrombin time, INR, platelet count, and significantly reduced blood product utilization and
fibrinogen level. These tests cannot assess the impact has led to improved outcomes.39 It has been shown
of the anticoagulant mechanisms outlined above2 to predict the need for massive transfusions during
and may overestimate the bleeding risk in cirrhosis. liver transplant, and transfusion protocols guided by
Prolongation of the prothrombin time and activated viscoelastic testing during liver transplant surgery
partial thromboplastin time indicates a decrease in have been shown to reduce the intraoperative use of
hepatic synthesis of procoagulation factors and cor- blood products without an associated increased rate
relates with hepatic function, but this does not ade- of bleeding.40,41
quately quantify bleeding risk. Outside the operating room, a small study by Chau
Of importance is that the INR is standardized et al42 found that abnormal results on thromboelas-
and validated using plasma from patients receiving tography were associated with risk of rebleeding in
vitamin K antagonists such as warfarin. There is patients with esophageal variceal hemorrhages, but
no standard reference plasma that could be used in data on predicting spontaneous bleeding risk remain
clinical practice to express a normalized ratio of the limited.
prothrombin time for patients with cirrhosis.15 Thromboelastography is highly reproducible and
Similarly, the quantitative decrease of platelet routinely shows normal coagulation profiles in patients
counts and fibrinogen levels in cirrhosis is balanced with cirrhosis in stable condition.43 For example, in
by qualitative changes in platelet activation and 273 patients with compensated cirrhosis, Stravitz44
528 CLEVELAND CLINIC JOURNAL OF MEDICINE VOLUME 89 • NUMBER 9 SEPTEMBER 2022

Downloaded from www.ccjm.org on April 19, 2024. For personal use only. All other uses require permission.
 SINGH AND COLLEAGUES

Tracing Pattern

Normal
R time, K time and alpha angle normal

Anticoagulant use or hemophilia

(factor deficiency)
R time and K time prolonged
Maximum amplitude, alpha angle decreased

Use of platelet blockers

(thrombocytopenia, thrombocytopathy)
R time normal, K time prolonged
Maximum amplitude decreased

Use of fibrinolytic drugs

(urokinase, streptokinase, tissue plasminogen
activator)
R time normal
Maximum amplitude shows continuous
decrease
Clot lysis at 30 minutes > 7.5%
Clot lysis at 60 minutes > 15.0%

Hypercoagulation
R time and K time decreased
Maximum amplitude and alpha angle increased

Disseminated intravascular coagulation

Stage 1: Hypercoagulable state with secondary

fibrinolysis

Stage 2: Hypocoagulable state

2022

Figure 4. Results of thromboelastography in various conditions.

found that the thromboelastography parameters were in a study by Tripodi et al.45

within the normal range, even though the prothrom- However, the coagulopathy of liver disease is
bin time and INR were prolonged. Similarly, only 14 dynamic, and thromboelastography displays a more
(27%) of 51 patients with stable cirrhosis had abnor- hypocoagulable profile with increasing severity of
mal clotting times on rotational thromboelastography liver disease as well as in the setting of acute decom-

CLEVELAND CLINIC JOURNAL OF MEDICINE VOLUME 89 • NUMBER 9 SEPTEMBER 2022 529

Downloaded from www.ccjm.org on April 19, 2024. For personal use only. All other uses require permission.
CIRRHOTIC COAGULOPATHY

TABLE 1
Bleeding risk associated with invasive procedures
High-risk procedures
All major surgeries (cardiac, intra-abdominal, orthopedic, brain, spine)
Intracranial pressure catheter insertion
Endoscopy: large polypectomy with endoscopic mucosal resection or submucosal resection

Moderate-risk procedures
Lumbar puncture
Percutaneous or transjugular liver biopsy
Transarterial or percutaneous therapies for hepatocellular carcinoma
Transjugular intrahepatic portosystemic shunt
Endoscopy for percutaneous gastrostomy placement, biliary sphincterotomy

Low-risk procedures
Paracentesis
Thoracocentesis
Dental extraction
Cardiac catheterization
Central line placement
Endoscopy for diagnosis, variceal band ligation, uncomplicated polypectomy

Note: Risk is calculated based on relative vascularity, expected vascular breech, and potential clinical consequences. Risk should always be
defined by the clinician performing the procedure.

Adapted from Intagliata NM, Argo CK, Stine JG, et al. Concepts and controversies in haemostasis and thrombosis associated with liver disease: Proceedings of
the 7th International Coagulation in Liver Disease Conference. Thromb Haemost 2018; 118(8):1491–1506. doi:10.1055/s-0038-1666861, reference 55.

pensation. De Pietri et al46 compared the coagulation Gastroenterology Association52 state that though vis-
profiles of 261 patients with decompensated cirrhosis coelastic tests are promising, their clinical utility in
and Model for End-Stage Liver Disease scores between predicting bleeding risk in patients with liver disease
15 and 40 with those of 40 healthy participants. The is yet to be firmly established. The current Society for
latency time between test initiation and clot forma- Critical Care Medicine guidelines53 recommend the
tion (R time) was prolonged in 41.5% of the patients, use of viscoelastic testing over the conventional tests
and the ultimate strength of the clot (maximal ampli- in patients with cirrhosis in the intensive care unit.
tude) was weaker in 79.3% patients with cirrhosis.46 The clinical applicability of these tests is detailed in
Further, Lloyd-Donald et al47 reported that 34 crit- the next section.
ically ill patients with Child-Pugh class C cirrhosis
had longer R times, weaker clot strength, and reduced ■ PROPHYLACTIC OPTIMIZATION
clot lysis compared with 157 healthy controls. This OF COAGULOPATHY
further supports the dynamicity of liver disease and
the impact of underlying cirrhosis on coagulability. Coagulopathy in cirrhosis is widely interpreted as a
However, certain limitations restrict widespread risk factor for bleeding after an invasive procedure.
clinical implementation of viscoelastic testing. The The need to minimize the risk of coagulopathy before
normal limits are not standardized, and clinical trials procedures is a common dilemma for practitioners.
have used different cutoff values as indications for The risk of bleeding is mainly determined by the
treatment.48–51 Moreover, the impact of concomitant type of procedure, the clinical scenario, comorbidi-
conditions such as sepsis or acute kidney injury on ties, use of ultrasonographic guidance, and operator
these results has not been studied.51 experience.54 Various procedures are classified as
Accordingly, the current guidelines from the high-, intermediate-, or low-risk (Table 1).2,55
American Association for the Study of Liver Dis- Traditional coagulation tests do not predict post-
eases11 and a technical review from the American procedural bleeding complications.56 A meta-analysis
530 CLEVELAND CLINIC JOURNAL OF MEDICINE VOLUME 89 • NUMBER 9 SEPTEMBER 2022

Downloaded from www.ccjm.org on April 19, 2024. For personal use only. All other uses require permission.
 SINGH AND COLLEAGUES

of 29 studies including 13,276 patients found that of 1.5 L, which would raise the portal pressure by
neither elevated INR (OR 1.52, 95% CI 0.99–2.33) approximately 15.5 mm Hg.64 This is important, since
nor thrombocytopenia (OR 1.24, 95% CI 0.55–2.77) an elevated hepatic venous pressure gradient (> 12
significantly increased the risk of bleeding in patients mm Hg) is associated with an increased risk of vari-
with cirrhosis.57 Moreover, the mean INR did not ceal hemorrhage.65,66
significantly differ between patients with bleeding Large-volume blood product transfusions aimed at
complications and those without. However, there was correcting an elevated INR can therefore translate to
significant heterogeneity (I2 = 51%) in the pooled increased bleeding complications. This is supported
results, likely attributable to differences in the sever- by a recent multicenter retrospective study, which
ity of thrombocytopenia in various studies. The risk found that transfusion of fresh frozen plasma to man-
of bleeding was associated with the type of invasive age acute variceal bleeding increased the risk of death
procedure, but not with the results of conventional within 42 days (OR 9.41, 95% CI 3.71–23.90).67
tests of coagulopathy.57 Notably, patients who received fresh frozen plasma
Paracentesis, the most commonly performed pro- had a higher INR at baseline evaluation, and the
cedure in patients with cirrhosis, is considered low- patients who had died by 42 days had received a
risk and can be done safely even if the results of con- median of 3 units of fresh frozen plasma, compared
ventional coagulation tests are abnormal. In a study with 0 units in those who were alive at 42 days.
of 1,100 therapeutic paracenteses in 628 patients, of Accordingly, the current recommendations advise
whom 513 had cirrhosis of the liver and in whom the against routine preprocedural correction of INR or
mean INR was 1.7, no patients received prophylactic thrombocytopenia in patients with cirrhosis, particu-
preprocedural correction of INR, and no significant larly for low-risk procedures.2,11,52,68
bleeding events (defined as bleeding requiring hos-
pitalization) were reported.58 Can viscoelastic testing reduce transfusions?
By comparison, a study of 2,740 percutaneous liver As reviewed, viscoelastic tests may more accurately
biopsies reported an increased risk of bleeding in patients assess the global coagulation status. Recent random-
with INR greater than 1.3 and platelet counts less than ized controlled trials have evaluated the impact of
60 × 109/L.59 This area clearly needs further study. thromboelastography-guided prophylactic transfusion
protocols compared with the standard of care for the
Transfusion may not reduce bleeding, use of blood products and bleeding complications for
and it has its own risks invasive procedures and in the setting of variceal and
Furthermore, no studies have shown that giving pro- nonvariceal gastrointestinal bleeding.48–60 In all the
phylactic transfusions of fresh frozen plasma to correct studies, thromboelastography-guided therapy signifi-
an elevated INR reduces the risk of procedure-related cantly reduced transfusion of blood products (fresh
bleeding in patients with cirrhosis. Also, in vitro frozen plasma and platelets) compared with the stan-
experiments have demonstrated that transfusion of dard of care, while the incidence of postprocedure-re-
fresh frozen plasma does not increase coagulation lated bleeding between the groups was similar.
potential in patients with cirrhosis, as it supplies However, several limitations need to be considered
both procoagulant and anticoagulant factors in equal when interpreting these findings. Most importantly,
amounts. As a result, the increase in plasma levels of in these trials, the standard of care aimed to “correct”
procoagulant factors may correct an elevated INR, the INR and platelet counts to arbitrary near-normal
but thrombin-generating potential does not change, thresholds. This is not in line with current restrictive
or may even decrease.60,61 recommendations for transfusion. Furthermore, trans-
The current standard of practice is prejudiced fusion thresholds in the thromboelastography-based
toward the hypocoagulable state of coagulopathy and protocols varied among trials, and there are currently no
disregards the risks associated with blood product uniform and well-established transfusion thresholds for
transfusion.36 Acutely ill patients with cirrhosis are viscoelastic tests.49,50,69 It remains unclear if a restrictive
at increased risk for transfusion-related lung injury transfusion strategy based on viscoelastic testing is supe-
and complications from transfusion-related circu- rior to a restrictive strategy based on conventional tests.
latory overload.62 In a small classic study, every 100 The small number of patients and the very low bleeding
mL of volume expansion increased the portal pressure rates observed in these trials further limit their generaliz-
by 1.4 cm H2O (1.03 mm Hg).63 It is estimated that ability, as they may therefore be underpowered to detect
lowering the INR from 2.0 to 1.5 requires transfusion true differences between the 2 strategies.
CLEVELAND CLINIC JOURNAL OF MEDICINE VOLUME 89 • NUMBER 9 SEPTEMBER 2022 531

Downloaded from www.ccjm.org on April 19, 2024. For personal use only. All other uses require permission.
CIRRHOTIC COAGULOPATHY

In sum, viscoelastic tests are promising tools to such thresholds to facilitate translation into general
both assess the coagulopathy of cirrhosis and guide clinical practice. ■
preprocedural management of hemostasis, but their
current use is limited by a lack of validated trans- ■ DISCLOSURES
fusion thresholds and limited clinical availability Dr. Lindenmeyer has disclosed authorship for Merck Manuals. The other
authors have disclosed no relevant financial relationships which, in the
outside of the operating room or research setting.2 context of their contributions, could be perceived as a potential conflict
Further large-scale studies are needed to establish of interest.

■ REFERENCES 17. Pant A, Kopec AK, Luyendyk JP. Role of the blood coagulation cas-
cade in hepatic fibrosis. Am J Physiol Gastrointest Liver Physiol 2018;
1. Tripodi A, Primignani M, Mannucci PM, Caldwell SH. Changing 315(2):G171–G176. doi:10.1152/ajpgi.00402.2017
concepts of cirrhotic coagulopathy. Am J Gastroenterol 2017; 18. Mitchell O, Feldman DM, Diakow M, Sigal SH. The pathophysiolo-
112(2):274–281. doi:10.1038/ajg.2016.498 gy of thrombocytopenia in chronic liver disease. Hepat Med 2016;
2. O'Leary JG, Greenberg CS, Patton HM, Caldwell SH. AGA clinical 8:39–50. doi:10.2147/HMER.S74612
practice update: coagulation in cirrhosis. Gastroenterology 2019; 19. Tripodi A, Primignani M, Chantarangkul V, et al. Thrombin genera-
157(1):34–43.e1. doi:10.1053/j.gastro.2019.03.070 tion in patients with cirrhosis: the role of platelets. Hepatology 2006;
3. Drolz A, Ferlitsch A, Fuhrmann V. Management of coagulopathy 44(2):440–445. doi:10.1002/hep.21266
during bleeding and invasive procedures in patients with liver fail- 20. Lisman T, Bongers TN, Adelmeijer J, et al. Elevated levels of von Will-
ure. Visc Med 2018; 34(4):254–258. doi:10.1159/000491106 ebrand Factor in cirrhosis support platelet adhesion despite reduced
4. Basili S, Raparelli V, Napoleone L, et al. Platelet count does not pre- functional capacity. Hepatology 2006; 44(1):53–61.
dict bleeding in cirrhotic patients: results from the PRO-LIVER study. doi:10.1002/hep.21231
Am J Gastroenterol 2018; 113(3):368–375. doi:10.1038/ajg.2017.457 21. Mannucci PM, Canciani MT, Forza I, Lussana F, Lattuada A, Rossi E.
5. Shah NL, Northup PG, Caldwell SH. A clinical survey of bleeding, Changes in health and disease of the metalloprotease that cleaves
thrombosis, and blood product use in decompensated cirrhosis von Willebrand factor. Blood 2001; 98(9):2730–2735.
patients. Ann Hepatol 2012; 11(5):686–690. pmid:22947530 doi:10.1182/blood.v98.9.2730
6. Budnick IM, Davis JPE, Sundararaghavan A, et al. Transfusion 22. Vlachogiannakos J, Binas J, Siakavellas S, et al. Platelet activation
with cryoprecipitate for very low fibrinogen levels does not affect and hypercoagulability in patients with early primary biliary chol-
bleeding or survival in critically ill cirrhosis patients. Thromb Haemost angitis compared with healthy controls. Ann Gastroenterol 2021;
2021; 121(10):1317–1325. doi:10.1055/a-1355-3716 34(2):229–234. doi:10.20524/aog.2021.0572
7. Ghabril M, Agarwal S, Lacerda M, Chalasani N, Kwo P, Tector AJ. Por- 23. Green D. Coagulation cascade. Hemodial Int 2006; 10(suppl 2):S2–S4.
tal vein thrombosis is a risk factor for poor early outcomes after liver doi:10.1111/j.1542-4758.2006.00119.x
transplantation: analysis of risk factors and outcomes for portal vein 24. Zermatten MG, Fraga M, Moradpour D, et al. Hemostatic alterations
thrombosis in waitlisted patients. Transplantation 2016; 100(1): in patients with cirrhosis: from primary hemostasis to fibrinolysis.
126–133. doi:10.1097/TP.0000000000000785 Hepatology 2020; 71(6):2135–2148. doi:10.1002/hep.31201
8. Zhang Y, Xu BY, Wang XB, et al. Prevalence and clinical significance 25. Green D. Overview of blood coagulation. Hemodial Int 2001;
of portal vein thrombosis in patients with cirrhosis and acute decom- 5(1):70–73. doi:10.1111/hdi.2001.5.1.70
pensation. Clin Gastroenterol Hepatol 2020; 18(11):2564–2572.e1. 26. Tripodi A, Primignani M, Chantarangkul V, et al. An imbalance of
doi:10.1016/j.cgh.2020.02.037 pro- vs anti-coagulation factors in plasma from patients with
9. Okuda K, Ohnishi K, Kimura K, et al. Incidence of portal vein throm- cirrhosis. Gastroenterology 2009; 137(6):2105–2111.
bosis in liver cirrhosis. An angiographic study in 708 patients. Gastro- doi:10.1053/j.gastro.2009.08.045
enterology 1985; 89(2):279–286. doi:10.1016/0016-5085(85)90327-0 27. Groeneveld D, Porte RJ, Lisman T. Thrombomodulin-modified throm-
10. Nonami T, Yokoyama I, Iwatsuki S, Starzl TE. The incidence of bin generation testing detects a hypercoagulable state in patients
portal vein thrombosis at liver transplantation. Hepatology 1992; with cirrhosis regardless of the exact experimental conditions.
16(5):1195–1198. pmid:1427658 Thromb Res 2014; 134(3):753–756.
11. Northup PG, Garcia-Pagan JC, Garcia-Tsao G, et al. Vascular liver dis- doi:10.1016/j.thromres.2014.07.010
orders, portal vein thrombosis, and procedural bleeding in patients 28. Lisman T, Leebeek FW, Mosnier LO, et al. Thrombin-activatable
with liver disease: 2020 practice guidance by the American Associa- fibrinolysis inhibitor deficiency in cirrhosis is not associated with in-
tion for the Study of Liver Diseases. Hepatology 2021; 73(1):366–413. creased plasma fibrinolysis. Gastroenterology 2001; 121(1):131–139.
doi:10.1002/hep.31646 doi:10.1053/gast.2001.25481
12. Ki M, Choi HY, Kim KA, Kim BH, Jang ES, Jeong SH. Incidence, preva- 29. Hugenholtz GC, Macrae F, Adelmeijer J, et al. Procoagulant changes
lence and complications of Budd-Chiari syndrome in South Korea: in fibrin clot structure in patients with cirrhosis are associated with
a nationwide, population-based study. Liver Int 2016; 36(7): oxidative modifications of fibrinogen. J Thromb Haemost 2016;
1067–1073. doi:10.1111/liv.13008 14(5):1054–1066. doi:10.1111/jth.13278
13. Søgaard KK, Horváth-Puhó E, Grønbaek H, Jepsen P, Vilstrup H, 30. Narvaiza MJ, Fernández J, Cuesta B, Páramo JA, Rocha E. Role of sial-
Sørensen HT. Risk of venous thromboembolism in patients with liver ic acid in acquired dysfibrinogenemia associated with liver cirrhosis.
disease: a nationwide population-based case-control study. Am J Ric Clin Lab 1986; 16(4):563–568. doi:10.1007/BF02886840
Gastroenterol 2009; 104(1):96–101. doi:10.1038/ajg.2008.34 31. Iba T, Levy JH. Sepsis-induced coagulopathy and disseminated
14. Ocak G, Vossen CY, Verduijn M, et al. Risk of venous thrombosis in intravascular coagulation. Anesthesiology 2020; 132(5):1238–1245.
patients with major illnesses: results from the MEGA study. J Thromb doi:10.1097/ALN.0000000000003122
Haemost 2013; 11(1):116–123. doi:10.1111/jth.12043 32. Esmon CT. The interactions between inflammation and coagulation.
15. Tripodi A, Mannucci PM. The coagulopathy of chronic liver disease. Br J Haematol 2005; 131(4):417–430.
N Engl J Med 2011; 365(2):147–156. doi:10.1056/NEJMra1011170 doi:10.1111/j.1365-2141.2005.05753.x
16. Stravitz RT. Thrombosis and coagulopathy in the liver transplant 33. Driever EG, Lisman T. Effects of Inflammation on hemostasis in acute-
candidate and recipient. Clin Liver Dis (Hoboken) 2017; 9(1):11–17. ly ill patients with liver disease. Semin Thromb Hemost 2022; 48(5):
doi:10.1002/cld.606 596–606. doi:10.1055/s-0042-1742438

532 CLEVELAND CLINIC JOURNAL OF MEDICINE VOLUME 89 • NUMBER 9 SEPTEMBER 2022

Downloaded from www.ccjm.org on April 19, 2024. For personal use only. All other uses require permission.
 SINGH AND COLLEAGUES

34. Blasi A, Calvo A, Prado V, et al. Coagulation failure in patients with 161(5):1630–1656. doi:10.1053/j.gastro.2021.09.004
acute-on-chronic liver failure and decompensated cirrhosis: beyond 53. Nanchal R, Subramanian R, Karvellas CJ, et al. Guidelines for the
the international normalized ratio. Hepatology 2018; 68(6): management of adult acute and acute-on-chronic liver failure in the
2325–2337. doi:10.1002/hep.30103 ICU: cardiovascular, endocrine, hematologic, pulmonary, and renal
35. Thakur M, Ahmed AB. A review of thromboelastography. Int J Periop considerations. Crit Care Med 2020; 48(3):e173–e191.
Ultrasound Appl Technol 2012; 1(1)25–29. doi:10.1097/CCM.0000000000004192
doi:10.5005/jp-journals-10027-1006 54. Schepis F, Turco L, Bianchini M, Villa E. Prevention and management
36. Singh AD, Shalimar. Use of blood products and drugs before of bleeding risk related to invasive procedures in cirrhosis. Semin
procedures in patients with cirrhosis. Clin Liver Dis (Hoboken) 2020; Liver Dis 2018; 38(3):215–229. doi:10.1055/s-0038-1660523
16(4):153–157. doi:10.1002/cld.906 55. Intagliata NM, Argo CK, Stine JG, et al. Concepts and controver-
37. Sakai T. Comparison between thromboelastography and throm- sies in haemostasis and thrombosis associated with liver disease:
boelastometry. Minerva Anestesiol 2019; 85(12):1346–1356. proceedings of the 7th International Coagulation in Liver Dis-
doi:10.23736/S0375-9393.19.13687-5 ease Conference. Thromb Haemost 2018; 118(8):1491–1506.
38. Whiting D, DiNardo JA. TEG and ROTEM: technology and clinical ap- doi:10.1055/s-0038-1666861
plications. Am J Hematol 2014; 89(2):228–232. doi:10.1002/ajh.23599 56. Janko N, Majeed A, Commins I, Kemp W, Roberts SK. Procedural
39. Serraino GF, Murphy GJ. Routine use of viscoelastic blood tests for bleeding risk, rather than conventional coagulation tests, predicts
diagnosis and treatment of coagulopathic bleeding in cardiac procedure related bleeding in cirrhosis. Eur J Gastroenterol Hepatol
surgery: updated systematic review and meta-analysis. Br J Anaesth 2022; 34(2):192–199. doi:10.1097/MEG.0000000000001948
2017; 118(6):823–833. doi:10.1093/bja/aex100 57. Kovalic AJ, Majeed CN, Samji NS, Thuluvath PJ, Satapathy SK. Sys-
40. Wang SC, Shieh JF, Chang KY, et al. Thromboelastography-guid- tematic review with meta-analysis: abnormalities in the international
ed transfusion decreases intraoperative blood transfusion during normalised ratio do not correlate with periprocedural bleeding
orthotopic liver transplantation: randomized clinical trial. Transplant events among patients with cirrhosis. Aliment Pharmacol Ther 2020;
Proc 2010; 42(7):2590–2593. doi:10.1016/j.transproceed.2010.05.144 52(8):1298–1310. doi:10.1111/apt.16078
41. Lawson PJ, Moore HB, Moore EE, et al. Preoperative thromboelas- 58. Grabau CM, Crago SF, Hoff LK, et al. Performance standards for ther-
tography maximum amplitude predicts massive transfusion in liver apeutic abdominal paracentesis. Hepatology 2004; 40(2):484–488.
transplantation. J Surg Res 2017; 220:171–175. doi:10.1002/hep.20317
doi:10.1016/j.jss.2017.05.115 59. Seeff LB, Everson GT, Morgan TR, et al; the HALT-C Trial Group. Com-
42. Chau TN, Chan YW, Patch D, Tokunaga S, Greenslade L, Burroughs AK. plication rate of percutaneous liver biopsies among persons with
Thrombelastographic changes and early rebleeding in cirrhotic advanced chronic liver disease in the HALT-C trial. Clin Gastroenterol
patients with variceal bleeding. Gut 1998; 43(2):267–271. Hepatol 2010; 8(10):877–883. doi:10.1016/j.cgh.2010.03.025
doi:10.1136/gut.43.2.267 60. Bernal W, Caldwell SH, Lisman T. Nails in the coffin of fresh frozen
43. George G, Manatasahit W, Balasubramanian M, Navarro V. Repro- plasma to prevent or treat bleeding in cirrhosis? J Hepatol 2020;
ducibility of TEG parameters in stable cirrhotics. Lab Med 2018; 72(1):12–13. doi:10.1016/j.jhep.2019.09.024
49(3):226–230. doi:10.1093/labmed/lmx041 61. Rassi AB, d'Amico EA, Tripodi A, et al. Fresh frozen plasma transfu-
44. Stravitz RT. Potential applications of thromboelastography in sion in patients with cirrhosis and coagulopathy: effect on conven-
patients with acute and chronic liver disease. Gastroenterol Hepatol tional coagulation tests and thrombomodulin-modified thrombin
(NY) 2012; 8(8):513–520. pmid:23293564 generation. J Hepatol 2020; 72(1):85–94.
45. Tripodi A, Primignani M, Chantarangkul V, et al. The coagulopathy doi:10.1016/j.jhep.2019.09.008
of cirrhosis assessed by thromboelastometry and its correlation with 62. Benson AB, Austin GL, Berg M, et al. Transfusion-related acute lung
conventional coagulation parameters. Thromb Res 2009; 124(1): injury in ICU patients admitted with gastrointestinal bleeding. Inten-
132–136. doi:10.1016/j.thromres.2008.11.008 sive Care Med 2010; 36(10):1710–1717.
46. De Pietri L, Bianchini M, Rompianesi G, Bertellini E, Begliomini B. doi:10.1007/s00134-010-1954-x
Thromboelastographic reference ranges for a cirrhotic patient pop- 63. Zimmon DS, Kessler RE. The portal pressure-blood volume relation-
ulation undergoing liver transplantation. World J Transplant 2016; ship in cirrhosis. Gut 1974; 15(2):99–101. doi:10.1136/gut.15.2.99
6(3):583–593. doi:10.5500/wjt.v6.i3.583 64. Giannini EG, Stravitz RT, Caldwell SH. Correction of hemostatic ab-
47. Lloyd-Donald P, Vasudevan A, Angus P, et al. Coagulation in acutely normalities and portal pressure variations in patients with cirrhosis.
ill patients with severe chronic liver disease: insights from throm- Hepatology 2014; 60(4):1442. doi:10.1002/hep.27029
boelastography. J Crit Care 2017; 38:215–224. 65. Garcia-Tsao G, Bosch J. Varices and variceal hemorrhage in cirrhosis:
doi:10.1016/j.jcrc.2016.10.030 a new view of an old problem. Clin Gastroenterol Hepatol 2015;
48. De Pietri L, Bianchini M, Montalti R, et al. Thrombelastography-guid- 13(12):2109–2117. doi:10.1016/j.cgh.2015.07.012
ed blood product use before invasive procedures in cirrhosis with 66. Kim JN, Sohn KM, Kim MY, et al. Relationship between the hepatic
severe coagulopathy: a randomized, controlled trial. Hepatology venous pressure gradient and first variceal hemorrhage in patients
2016; 63(2):566–573. doi:10.1002/hep.28148 with cirrhosis: a multicenter retrospective study in Korea. Clin Mol
49. Vuyyuru SK, Singh AD, Gamanagatti SR, Rout G, Gunjan D, Shalimar. Hepatol 2012; 18(4):391–396. doi:10.3350/cmh.2012.18.4.391
A randomized control trial of thromboelastography-guided transfu- 67. Mohanty A, Kapuria D, Canakis A, et al. Fresh frozen plasma transfu-
sion in cirrhosis for high-risk invasive liver-related procedures. Dig Dis sion in acute variceal haemorrhage: results from a multicentre cohort
Sci 2020; 65(7):2104–2111. doi:10.1007/s10620-019-05939-2 study. Liver Int 2021; 41(8):1901–1908. doi:10.1111/liv.14936
50. Rout G, Shalimar, Gunjan D, et al. Thromboelastography-guid- 68. European Association for the Study of the Liver. EASL Clinical
ed blood product transfusion in cirrhosis patients with variceal Practice Guidelines: vascular diseases of the liver. J Hepatol 2016;
bleeding: a randomized controlled trial. J Clin Gastroenterol 2020; 64(1):179–202. doi:10.1016/j.jhep.2015.07.040
54(3):255–262. doi:10.1097/MCG.0000000000001214 69. Kumar M, Ahmad J, Maiwall R, et al. Thromboelastography-guided
51. Hunt H, Stanworth S, Curry N, et al. Thromboelastography (TEG) and blood component use in patients with cirrhosis with nonvariceal
rotational thromboelastometry (ROTEM) for trauma induced coagu- bleeding: a randomized controlled trial. Hepatology 2020; 71(1):
lopathy in adult trauma patients with bleeding. Cochrane Database 235–246. doi:10.1002/hep.30794
Syst Rev 2015; 2015(2):CD010438.
doi:10.1002/14651858.CD010438.pub2 Address: Christina C. Lindenmeyer, MD, Department of Gastroenterology
52. Intagliata NM, Davitkov P, Allen AM, Falck-Ytter YT, Stine JG. AGA and Hepatology, A30, Cleveland Clinic, 9500 Euclid Avenue, Cleveland,
Technical review on coagulation in cirrhosis. Gastroenterology 2021; OH 44195; lindenc@ccf.org

CLEVELAND CLINIC JOURNAL OF MEDICINE VOLUME 89 • NUMBER 9 SEPTEMBER 2022 533

Downloaded from www.ccjm.org on April 19, 2024. For personal use only. All other uses require permission.