ORIGINAL ARTICLE
Clot wave analysis and thromboembolic score in liver cirrhosis:
two opposing phenomena
M. F. RUBERTO*, O. SORBELLO † , A. CIVOLANI † , D. BARCELLONA*
*Internal Medicine and
Haemocoagulophaties Unit,
University of Cagliari, Cagliari,
Italy
†
Gastroenterology Unit,
Policlinico Universitario ‘Duilio
Casula’, University of Cagliari,
Cagliari, Italy
Correspondence:
Prof Francesco Marongiu, Inter-
nal Medicine and Haemostasis
and Thrombosis Unit, Policlinico
Universitario ‘Duilio Casula’,
University of Cagliari, SS 554,
Km 4,500, 09042 Monserrato
(Cagliari), Cagliari, Italy.
Tel.: +390706754188;
E-mail:
marongiu@medicina.unica.it
doi:10.1111/ijlh.12635
Received 26 July 2016; accepted
for publication 7 January 2017
Keywords
Clot waveform analysis,
activated partial thromboplastin
time derivatives, liver cirrhosis,
venous thrombosis,
hemorrhage, risk assessment
INTRODUCTION
It has been reported that the
tests, prothrombin time (PT),
thromboplastin time (aPTT) are
patients to express their risk of
© 2017 John Wiley & Sons Ltd, Int. Jnl. Lab. Hem. 2017, 39, 369–374
INTERNAT IONAL JOURNAL OF LABORATO RY HEMATO LOGY
, L. DEMELIA † , F. MARONGIU*
S U M M A RY
Introduction: Prothrombin time (PT) and activated partial thrombo-
plastin time (aPTT) are abnormal but unreliable in cirrhotic patients
to express their risk of bleeding. However, these patients may also
suffer from thrombotic episodes. In order to investigate the dynam-
ics of the formation of fibrin, the clot waveform analysis (CWA) of
aPTT was studied together with a score for the evaluation of the
thromboembolic risk.
Methods: CWA in terms of velocity (1st derivative), acceleration
(2nd derivatives) and density (Delta) of aPTT and the Padua Predic-
tion Score (PPS) for venous thromboembolism were studied in 191
cirrhotic patients.
Results: CWA values were lower in the cirrhotic patients when
compared to the control groups. However, Delta, 1st and 2nd
derivatives were higher in cirrhotic patients with elevated PPS in
comparison to those with a low PPS. The 1st derivative was signifi-
cantly associated with a high PPS score (>4): OR: 2.66, CI: 95%
1.23–5.78.
Conclusions: Two opposing tendencies seem to be present in cirrhotic
disease: the first shows a weakness of clot formation while the sec-
ond a predisposition towards thrombosis, identified by the PPS.
words, the concept that cirrhotic patients are auto-
anticoagulated has been revised because these patients
coagulative screening have shown a high thromboembolic risk [2] and their
and activated partial thrombin generation is similar to normal subjects
unreliable in cirrhotic when this determination is carried out by adding
bleeding [1]. In other thrombomodulin, a natural anticoagulant [3].
369
370 M. F. RUBERTO ET AL. | CLOT WAVE ANALYSIS AND LIVER CIRRHOSIS

However, there is no doubt that these tests are abnor- imaging, and histological criteria. Characteristics of
mal in liver cirrhosis, expressing a decrease in overall the patients studied are presented in Table 1. Further,
hepatic synthesis of coagulation factors [4]. In particu- two control groups of 100 normal subjects (55 males
lar, a prolonged aPTT is typical of a deficiency of fac- and 45 females, median age 60, range 40–85 years)
tors XII, XI, X, IX, VIII, V, II, fibrinogen but also of and 282 patients (127 males and 155 females, median
the presence of lupus anticoagulant or inhibitors of age 64, range 40–80 years) who were consecutively
individual clotting factors. Nevertheless, a long aPTT admitted to the clinical ward of the Internal Medicine
can have clinical significance, especially as regards and Haemostasis and Thrombosis Unit of the Univer-
bleeding risk, for example, in congenital or acquired sity Hospital of Monserrato were also studied.
hemophilia [5]. On the other hand, a short aPTT may Informed consent was obtained from all patients and
be a manifestation of a hypercoagulable state as controls.
reported by Tripodi et al. [6]. A Padua Prediction Score (PPS) [9] was determined
To study the dynamics of clot formation in patients for each cirrhotic patient. PPS has been recommended
with liver cirrhosis, we analyzed the behavior of aPTT by the American College of Chest Physicians [10] to
that, unlike PT, investigates all coagulation factors evaluate the thromboembolic risk in internal medicine
except VII. What is not fully known is whether the patients by assigning points to 11 common VTE risk
abnormalities of aPTT in patients with chronic liver factors. A PPS score ≥4 indicated a high thromboem-
disease are associated with a reduction in the bolic risk. Clinical history positive for hemorrhagic
‘strength’ of the clot. To investigate the dynamics of events was also recorded for each patient. As the pur-
the formation of fibrin, the clot waveform analysis pose of this study was to obtain data on CWA related
(CWA) of aPTT was studied in terms of velocity (1st to aPTT requested for routine evaluation of the
derivative), acceleration (2nd derivative), and clot patients, our study did not require ethical considera-
density (Delta). These characteristics of the clot are tions.
available today by means of an automated coagulome-
ter through the use of an implemented software [7].
METHODS
This topic has been the subject of an official commu-
nication by the Subcommittees of the International Blood sampling was performed using Na-citrate 3.2%
Society Thrombosis Haemostasis (ISTH) aimed at the (0.109 M) vacuum bottles (Venosafe; Terumo Europe,
standardization of CWA and the recommendations for 3.2%). Samples were centrifuged at 1500 g for 20 min
its clinical application and research [8]. We analyzed at room temperature. Plasma was separated, and a sil-
these ‘hidden’ parameters of aPTT, that is, the CWA ica liquid aPTT ratio and the CWA were immediately
in a group of cirrhotic patients and in two control carried out for each patients and controls. All analyses
groups formed by normal subjects and patients admit-
ted to the clinical ward of an Internal Medicine Unit.
Finally, a score for the evaluation of a thromboem- Table 1. Types of liver cirrhosis involved in the study
bolic risk was also computed to investigate on a possi-
Types of liver cirrhosis n
ble relationship between CWA and a predisposition
toward thrombosis. Alcoholic 38
Alcoholic HCV related 7
Alcoholic HBV related 5
PAT I E N T S HCV related 102
HBV related 17
We studied 191 consecutive cirrhotic patients (137 HBV-HDV 1
males and 54 females, median age 64, range 40– Primary Biliary 11
87 years) referred to the outpatient unit or to the Autoimmune 3
clinical ward of the Gastroenterology Unit of the HCV Autoimmune 1
University Hospital of Monserrato (Cagliari, Italy) Alcoholic Autoimmune 2
Wilson disease 1
from December 2014 to November 2015. Diagnosis of
Cryptogenetic 3
cirrhosis was made based on clinical, laboratory,

© 2017 John Wiley & Sons Ltd, Int. Jnl. Lab. Hem. 2017, 39, 369–374

were performed using an ACL TOP 500 CTS Family
(Werfen, Barcelona, Spain) which uses a turbidimetric
method for clot detection.
In this type of coagulometer, 0% absorbance
defines the precoagulation phase and then the absor-
bance (mAbs) increases after the initiation of clotting.
The parameters examined inside the clot waveform
were as follows: (i) time at which maximum velocity
of clot formation was reached (1st derivative,
expressed as mAbs/s1), (ii) time at which maximum
change in velocity (maximum acceleration) of clot
formation was reached (2nd derivative, expressed as
mAbs/s2), and (iii) total change in optical density, ref-
erenced against a calibration curve, that is, the maxi-
mum density of the clot (Delta, expressed as mAbs).
Coefficient of variation (CV %) achieved by testing
over different runs the same plasma of different types
of samples (normal, cirrhotic, and internistic patient)
is shown in Table 2.
Statistical analysis
As the variables examined were not normally dis-
tributed, data are presented as median and ranges.
Accordingly, the Mann–Whitney test for independent
data, the nonparametric analysis of variance (Kruskal–
Wallis), and the post hoc Dunn’s test were used for sta-
tistical evaluation of the data. The v2 test was also
used for proportions comparisons. Multivariate and
stepwise logistic regressions were performed to
exclude independent variables (aPTT ratio, and 1st
and 2nd derivatives) that were not associated with
the dependent variable (thromboembolic score as
dichotomous, that is, 0 or 1 if the score was <4 or ≥4,
respectively). Logistic regression analysis was per-
formed through a backward stepwise procedure by
eliminating firstly the least significant interactions and
Table 2. Coefficients of variation (CV %) of clot
wave analysis related to different types of patients
CV% CV% normal
Parameter cirrhotics subjects
1st derivative 4.35 3.18
2nd derivative 4.56 4.40
Delta 6.90 5.50
© 2017 John Wiley & Sons Ltd, Int. Jnl. Lab. Hem. 2017, 39, 369–374
M. F. RUBERTO ET AL. | CLOT WAVE ANALYSIS AND LIVER CIRRHOSIS 371
then the insignificant independent variables. A value
between 0 and 1 was used to identify a value immedi-
ately superior or not to the lowest value of the nor-
mal range of the independent variables. The results
are reported indicating the odds ratio with a 95%
confidence interval. MEDCALC software (version 15.1,
Ostend, Belgium) was used for the statistical analysis
of the data.
R E S U LT S
A high percentage of cirrhotic patients (48 of 191,
25%) showed a high thromboembolic risk while the
clinical history in these patients for venous thrombosis
was positive in 6 of 48 patients (five portal vein
thrombosis and one splenic and mesenteric vein
thrombosis) (12.5%). Only one patient in the group
with a low thromboembolic risk (0.7%) had a clinical
history positive for venous portal vein thrombosis (v2:
11.02, P < 0.0009). Clinical history showed positive
bleeding episodes in 74 of 191 patients (39%). The
bleeding episodes were as follows: esophageal varices
(n = 13), epistaxis (n = 35), cerebral (n = 1), gastric
and duodenal angiodysplasia (n = 3), postsurgery
(n = 1), peptic ulcer (n = 3), macroscopic hematuria
(n = 1), gum bleeding (n = 1), rectorrhagia (n = 2),
and hemorrhoids (n = 14).
Laboratory investigation showed a significant
increase in aPTT ratio values in the cirrhotic group
when compared to the normal subjects and internal
medicine patients, whereas a significant decrease in
the values of 1st derivative and 2nd derivative was
found in cirrhotic patients when compared to the con-
trol groups. All these parameters were significantly
higher in the internal medicine patients when com-
pared to normal subjects (Table 3).
These parameters were found to be higher in the
group of cirrhotic patients with a high PPS in compar-
ison with those with a low PPS, even if lower
(P < 0.05) than those found in the normal subjects.
Only the aPTT ratio was similar in both groups of cir-
rhotic patients (Table 4).
CV%
When patients with a positive history of bleeding
internistic
patients were examined, different values for these parameters
were observed as follows: 1st derivative, 2nd deriva-
1.76 tive, and Delta were significantly lower in those
3.78
patients who presented a clinical history of bleeding
6.53
in comparison with patients who did not (Table 5).
372 M. F. RUBERTO ET AL. | CLOT WAVE ANALYSIS AND LIVER CIRRHOSIS

Table 3. All parameters studied in cirrhotic patients and controls. Data are expressed as median and range

Internal medicine
Parameter Cirrhotics (n = 191) Normal subjects (n = 100) patients (n = 282)

aPTT Ratio 1.14, 0.77–1.70* 1.06, 0.80–1.27 1.04, 0.66–2.14

1st derivative mAbs/s 135, 23–410** 188, 115–339** 236, 67–652
2nd derivative mAbs/s2 391, 21–1422** 577, 257–1082** 747, 113–2180
Delta mAbs 57, 7–193** 76, 40–162 94, 29–336

Dunn’s test after Kruskal–Wallis analysis of variance (P < 0.0001).

*P < 0.05 vs normal subjects and internal medicine patients; normal subjects vs internal medicine patients ns.
**P < 0.05 vs normal subjects and internal medicine patients; normal subjects vs internal medicine patients P < 0.05.

Table 4. All parameters studied in cirrhotic patients with high and low thromboembolic score. Data are expressed
as median and range

Thromboembolic score ≥4 Thromboembolic score <4

Parameter = 48) = 143) P

aPTT Ratio 1.12, 0.77–1.70 1.15, 0.79–1.70 0.088

1st derivative mAbs/s 154, 40–398 126, 23–410 0.010
2nd derivative mAbs/s2 500, 102–1317 379, 21–1422 0.020
Delta mAbs 64, 19–193 55, 7–163 0.020

Table 5. All parameters studied in cirrhotic patients with and without bleeding. Data are expressed as median and
range

Parameter With bleeding (n = 74) Without bleeding (n = 117) P

aPTT Ratio 1.15, 0.77–1.70 1.14, 0.79–1.70 0.427

1st derivative mAbs/s 119, 29–352 142, 23–410 0.025
2nd derivative mAbs/s2 369, 51–1067 410, 21–1422 0.075
Delta mAbs 52, 16–185 60, 7–193 0.012

Only the aPTT ratio was similar between these groups,
as we found in patients divided on the basis of the
PPS score. The final logistic regression model included
four variables: aPTT ratio, 1st derivative, 2nd deriva-
tive, and Delta. Only the 1st derivative was signifi-
cantly associated with the PPS ≥4 (OR: 2.67, CI: 95%
1.23–5.78). The stepwise procedure excluded the
other variables. The Hosmer and Lemeshow test
(P = 1.0) indicates a good logistic regression model fit.
No interactions among the variables were found to be
significantly associated with the PPS ≥4.
DISCUSSION
The CWA has recently been standardized by ISTH
[8] recommending the use of a coagulometer, such
as the ACL TOP employed in this study, which has
a software capable of expressing absorbance as the
whole process of clot formation. On the other hand,
the study of the derivatives has been used for the
evaluation of clot formation in bleeding disorders
such as hemophilia, showing a relationship between
very low levels of factor VIII and 2nd derivative of
aPTT [11]. The authors found that CWA could be
useful for the investigation of the clinical phenotype
of individual patients. We chose to study the CWA
to obtain specific information on the clot formation
of aPTT, such as velocity (1st derivative), accelera-
tion (2nd derivative), and clot density in order to
assess whether these variables were altered in
patients with liver cirrhosis as compared to normal
subjects and to patients with different clinical

© 2017 John Wiley & Sons Ltd, Int. Jnl. Lab. Hem. 2017, 39, 369–374

conditions such as those admitted to a clinical ward
of an Internal Medicine Unit.
Results show a large overlap of the data among the
groups but indicate that clot formation is weaker in
cirrhotic patients when compared to healthy controls
that, in turn, show lower values than those found in
the group of internal medicine patients. The latter
comprised patients with several comorbidities who are
associated with an increased activation of blood coag-
ulation as was demonstrated by the thrombin genera-
tion test [12] and thus to an increased thrombotic risk
[13]. A graphical example of the different curves
obtained in the three groups of subjects is shown in
Figure 1 showing large differences in mAbsorbances.
However, if in the cirrhotic group as a whole we
found a decrease in the parameters underlying clot
formation, in those groups with a high PPS score, an
increase in velocity, acceleration, and clot density was
found. The differences between cirrhotic patients with
or without high thromboembolic score seem to indi-
cate that in these patients, a relationship between a
clinical prothrombotic phenotype and biochemical
hypercoagulability seems to exist. The same, but
inverse, conceptual approach could be applicable to
cirrhotic patients with or without a positive clinical
history for bleeding. In these patients, we found a sig-
nificant difference in CWA because it showed lower
values in those with a positive clinical history for
bleeding, even if the events were of minor entity. In
other words, the different types of examined subjects
appear to be stratified on the basis of 1st, 2nd deriva-
tives, and Delta.
Our study also indicates how routine tests such as
aPTT are inadequate to represent the real drawback of
clot formation that we have studied with CWA. In
fact, even if the aPTT ratio is higher in cirrhotic
(a)
Figure 1. Examples of Delta, 1st derivative, and 2nd derivative in a cirrhotic patient (a), a normal subject (b), and
an internal medicine patient. Large differences in mAbsorbances among the three subjects are evident.
© 2017 John Wiley & Sons Ltd, Int. Jnl. Lab. Hem. 2017, 39, 369–374
M. F. RUBERTO ET AL. | CLOT WAVE ANALYSIS AND LIVER CIRRHOSIS 373
patients, it does not distinguish between those who
have a high or low thromboembolic score and those
who have or do not have a history of bleeding. Again,
the aPTT ratio was shown to be similar between nor-
mal subjects and patients hospitalized in a clinical
ward of Internal Medicine, but this study demon-
strates that the density of the CWA, the velocity, and
the acceleration of clot formation are significantly
higher in this group of patients. However, the 1st
derivative only showed a significant association with
PPS. Thus, from a practical point of view, both PPS
and the 1st derivative may be useful in the evaluation
of the thromboembolic risk in the single cirrhotic
patient. On the other hand, it is known that cirrhotic
patient can present both bleeding and thrombosis as
reported by Rodrıguez-Castro et al. [14]. Our study
could be interpreted as a confirm of this conceptual
approach even if from a different point of view.
We recognize some limitations: first, this study is
retrospective as for previous thrombotic and bleeding
episodes. However, this retrospective way of analysis
was chosen by us as we dealt with new aspects of a
pathological condition for the first time. We wished to
know whether this study could be of value in plan-
ning a prospective evaluation of possible predictive
values of both derivatives and PPS. Results of this
study seem to support our choice. Second, PPS has
not been clinically validated in cirrhosis. However,
patients with a high PPS score suffered from venous
thrombosis much more than those with a low score
indicate that this easy way of detecting a thrombotic
predisposition is of value even in liver cirrhosis. So, it
could be used in every patient admitted to a clinical
ward, including those with liver cirrhosis because this
clinical condition is frequently found in the daily
practice. The fact that in our study, a significant
(b) (c)
374 M. F. RUBERTO ET AL. | CLOT WAVE ANALYSIS AND LIVER CIRRHOSIS

number of patients with a high PPS score had a posi-
tive history of venous thrombosis, typical of cirrhotic
patients, means that PPS score may have a role in the
evaluation of the propensity of these patients to
develop thrombosis.
Third, the sample size is relatively small and we
were unable to carry out other biochemical evalua-
tions of hypo- or hypercoagulable states to validate
these findings. However, to the best of our knowl-
edge, this study shows for the first time that two
opposing tendencies seem to be present in cirrhotic
disease: The first shows toward a weakness of clot for-
mation, whereas the second shows a predisposition
toward thrombosis, identified by the PPS, that was
associated with an increased number of thrombotic
events. The latter should be always evaluated in
patients with liver cirrhosis.
CONFLICT OF INTEREST
The authors declare no conflict of interest.
AC K N OW L E D G E M E N T
We thank Mr Barry Mark Wheaton for his editing
assistance.

REFERENCES thromboembolism. Blood 2004;104: Chest Physicians Evidence-Based Clinical

1. Tripodi A. Liver Disease and Hemostatic
(Dys)function. Semin Thromb Hemost
2015;41:462–7.
2. Buresi M, Hull R, Coffin CS. Venous
thromboembolism in cirrhosis: a review of
the literature. Can J Gastroenterol
2012;26:905–8.
3. Tripodi A, Salerno F, Chantarangkul V,
Clerici M, Cazzaniga M, Primignani M,
Mannucci PM. Evidence of normal throm-
bin generation in cirrhosis despite abnor-
mal conventional coagulation tests.
Hepatology 2005;4:553–8.
4. Amitrano L, Guardascione MA, Brancaccio
V, Balzano A. Coagulation disorders in
liver disease. Semin Liver Dis 2002;22:83–
96.
5. White GC 2nd. The partial thromboplastin
time: defining an era in coagulation. J
Thromb Haemost 2003;1: 2267–70.
6. Tripodi A, Chantarangkul V, Martinelli I,
Bucciarelli P, Mannucci PM. A shortened
activated partial thromboplastin time is
associated with the risk of venous
3631–4.
7. Solano C, Zerafa P, Bird R. A study of atyp-
ical APTT derivative curves on the ACL
TOP coagulation analyser. Int J Lab Hema-
tol 2011;33:67–78.
8. Shima M, Thachil J, Nair SC, Srivastava A,
Scientific and Standardization Committee.
Towards standardization of clot waveform
analysis and recommendations for its clini-
cal applications. J Thromb Haemost
2013;11:1417–20.
9. Barbar S, Noventa F, Rossetto V, Ferrari A,
Brandolin B, Perlati M, De Bon E, Tormene
D, Pagnan A, Prandoni P. A risk assessment
model for the identification of hospitalized
medical patients at risk for venous throm-
boembolism: the Padua Prediction Score. J
Thromb Haemost 2010;8:2450–7.
10. Kahn SR, Lim W, Dunn AS, Cushman M,
Dentali F, Akl EA, Cook DJ, Balekian AA,
Klein RC, Le H, Schulman S, Murad MH;
American College of Chest Physicians.
Prevention of VTE in nonsurgical patients:
antithrombotic Therapy and Prevention of
Thrombosis, 9th edn. American College of
Practice Guidelines. Chest 2012;2(Suppl):
e195S–226S.
11. Shima M, Matsumoto T, Fukuda K, Kubota
Y, Tanaka I, Nishiya K, Giles AR, Yoshioka
A. The utility of activated partial thrombo-
plastin time (aPTT) clot waveform analysis
in the investigation of hemophilia A
patients with very low levels of factor VIII
activity (FVIII:C). Thromb Haemost
2002;87:436–41.
12. Espitia O, Fouassier M, Hardouin JB, Pisto-
rius MA, Agard C, Planchon B, Trossaert M,
Pottier P. Thrombin generation assay in hos-
pitalized nonsurgical patients: a new tool to
assess venous thromboembolism risk? Clin
Appl Thromb Hemost 2017;23:45–51.
13. Francis CW. Clinical practice. Prophylaxis
for thromboembolism in hospitalized medi-
cal patients. N Engl J Med 2007;356:1438–
44.
14. Rodrıguez-Castro KI, Antonello A, Fer-
rarese A. Spontaneous bleeding or throm-
bosis in cirrhosis: what should be feared
the most? World J Hepatol 2015;7:1818–
27.

© 2017 John Wiley & Sons Ltd, Int. Jnl. Lab. Hem. 2017, 39, 369–374