Вы находитесь на странице: 1из 8


Received: 28 April 2017    Accepted: 11 July 2017

DOI: 10.1111/ijlh.12724


Clot waveform analysis: Where do we stand in 2017?

P. O. Sevenet  | F. Depasse

Diagnostica Stago S.A.S, Asnières sur Seine,

France Abstract
Analysis of the optical waveform generated during global coagulation assays, such as
Pierre-Olivier Sevenet, Diagnostica Stago, activated partial thromboplastin time and prothrombin time, can provide much pre-
Clinical Development, Asnières-sur-Seine, cious information on the global coagulation state of the plasma sample tested, in addi-
Email: pierreolivier.sevenet@stago.com tion to a single clotting time. Many studies have been published concerning patient
diagnosis and management in haemophilia A, and in the early diagnosis and prognosis
of disseminated intravascular coagulation and sepsis. However, many other works
have also been published on further potential clinical applications such as lupus anti-
coagulant diagnosis and anticoagulant monitoring. Altogether, these publications have
demonstrated the ability for clot waveform analysis (CWA) to improve patient man-
agement, especially as this tool is inexpensive, rapid and readily available on coagula-
tion analysers with optical detection systems. By an extensive review of the literature
related to studies performed on CWA, this publication aims at providing a review of
current knowledge in this specific field, ranging from research data to potential clinical
applications and future trends.

activated partial thromboplastin time, biphasic waveform, clot waveform, disseminated
intravascular coagulation, haemophilia, transmittance

1 |  INTRODUCTION Both quantitative and qualitative CWA parameters have been
shown to be associated with pathophysiological processes and have
Coagulation is a highly complex physiological mechanism that involves a potential use in clinical applications. Among them, the clinical phe-
multiple plasma proteins, cellular components and the vessel wall. notype of severe haemophilia A (HA) patients could very accurately
Increasing coagulation and fibrinolysis knowledge has led to major be predicted by CWA.1,2 The management of haemophilia treatments
improvements in clinical practice decisions, resulting in optimized and and bypass therapy in HA patients with inhibitors is also of great in-
safer patient management in both haemorrhagic and thrombotic set- terest.3,4 Additionally, it has been shown that CWA can help in the
tings. Among the numerous laboratory assays exploring haemostasis, early diagnosis and prognosis of sepsis and disseminated intravascular
global functional assays have been generating more and more inter- coagulation (DIC).5,6
est over time. By analysing an end-­point of the coagulation cascade This review has been written using an exhaustive analysis of orig-
(eg thrombin or fibrin clot), they combine, within a single assay, the inal articles in English, present in the MEDLINE/PubMed database,
analysis of a complete process and allow the detection of several co- about CWA. Fifty-­six publications concerning the use of CWA were
agulation abnormalities. The most widely known global coagulation found using the search fields: “(waveform[all fields] AND (clot[all
assays are the thrombin generation test (TGT) and the viscoelastomet- fields] OR activated partial thromboplastin time[all fields])) AND (anal-
rics assays (such as ROTEM® or TEG®). Through the evaluation of the ysis[all fields] OR biphasic[all fields] OR haemophilia[all fields] OR dis-
kinetic of fibrin formation during tests such as activated partial throm- seminated intravascular coagulation[all fields])” and “activated partial
boplastin time (aPTT) or prothrombin time (PT), clot waveform analysis thromboplastin time[all fields] AND (reaction curve[all fields] OR op-
(CWA) is also considered as a global coagulation assay. tical profil[all fields] OR transmittance[all fields] OR first derivative[all

Int J Lab Hem. 2017;39:561–568. © 2017 John Wiley & Sons Ltd |  561
562       SEVENET and DEPASSE

fields] OR second derivative[all fields]).” After exclusion of noneligi- are 2 parameters of interest, frequently reported in the literature.
ble papers (reviews, editorials), 44 original articles were included for Denomination of these parameters varies depending on whether the
analysis. recorded signal is transmittance or absorbance. In the case of trans-
mittance, which is the most widely reported detection system in the
literature, maximum velocity is called “Min1,” maximum acceleration is
2 |  WHAT IS CLOT WAVEFORM ANALYSIS? called “Min2,” and maximum deceleration is called “Max2.” The normal
transmittance waveform pattern is illustrated in Figure 1. The pattern
Clot waveform analysis corresponds to the extended study of the of the absorbance detection system is the reverse of the transmittance
slope generated by an optical detection system during routine co- pattern. Hence, maximum velocity, maximum acceleration and maxi-
agulation assays, such as aPTT or PT. The optical detection system mum deceleration are called “Max1,” “Max2” and “Min2,” respectively.
recognizes the clot formation process by measuring changes in To aid comprehension, transmittance nomenclature will be used in this
transmittance, or absorbance, of a light beam through the analysed review.
sample. Transmittance or absorbance is continuously recorded over Besides quantitative parameters reflecting fibrin formation
time. The qualitative examination of the clot formation curve and the kinetics, CWA is able to identify other potential reactions of interest
multiple quantitative parameters provided by CWA can give valuable in clinical practice such as detection of the CRP-­VLDL complexes. The
information in addition to clotting time obtained by routine assays. formation of these complexes results in the biphasic waveform pattern
CWA is mainly based on the aPTT assay; however, PT or modified (BWP).6 This is of special relevance in the prediction of impending DIC
assays have also been used in various studies. The resultant curves as we will see further in this article.
obtained are called “waveform” because of their sigmoidal profile To carry out CWA, an automated coagulation analyser that uses
(Figure 1).7 an optical-­based system for clot formation detection is required. The
The coagulation process can be divided into 3 distinct periods, the analyser is usually equipped with software that automatically comput-
precoagulation phase, the coagulation phase and the postcoagulation erizes absorption or transmittance raw data and translates them into
phase. These 3 phases can be characterized by parameters defining CWA parameters. Alternatively, raw data can also be exported and
the time interval, the rate, the “slope” and the magnitude of the sig- processed by external software.
nal variation during the reaction. Multiple quantitative parameters can The standardization of CWA methods was proposed during
be recorded in addition to clotting time. While the transmittance (or the Scientific and Standardization Committee (SSC) sessions of the
absorbance) curve indicates the appearance of the fibrin clot in the International Society of Haemostasis and Thrombosis (ISTH) in 2013
sample, the first derivative of the curve reflects the velocity of the by Shima et al8 They formulated the minimal requirements for per-
clot formation. The maximum absolute value of the first derivative is forming CWA, which are the use of colourless and clear reagents, a
thus considered as the maximum velocity. The second derivative rep- system able to detect sensitive changes in the opacity of the sample,
resents the acceleration of clot formation. The maximum acceleration and an adjusted aPTT to ensure a normal clotting time between 30
and maximum deceleration (steepest point of the deceleration slope) and 40 seconds.

F I G U R E   1   Representation of a normal
clot waveform transmittance pattern,
with first and second derivatives. Min1
corresponds to the maximum velocity of
the clotting process, Min2 corresponds
to the maximum acceleration, and Max2
corresponds to the maximum deceleration

3 | POTENTIAL CLINICAL INDICATIONS is also supported by the publication of Yada and colleagues, who in-
vestigated a specific variant (Arg1781His of F8 gene) associated with
3.1 | Haemophilia moderate clinical phenotype in severe haemophilic patients. CWA, as
well as thrombin generation and thromboelastography, shows that re-
The first experiments with CWA in haemophilia were performed in sults from these severe patients are more comparable to plasma at
1997 by Braun et al, who suggested the potential expanded interpre- 5 IU/dL of FVIII:C than <1.0 IU/dL. Even though the mechanism re-
tation of aPTT and PT assays using optical transmittance data.7 Then, mains only partially understood, it would appear that binding affinity
Shima et al also demonstrated the promising correlation between between FVIII and FX is higher in Arg1781His patients than in normal
small amount of factor VIII (FVIII) in the sample and coinciding vari- FVIII patients.12
ation of CWA parameters. They observed significant qualitative dif- The absolute maximum values of the first and second derivatives
ferences in aPTT clot waveform curves across 36 patients with severe of the aPTT curve (Min1 and Min2, respectively) are the most widely
HA (ie patients with FVIII one-­stage clotting assay result <1.0 IU/dL). used parameters for CWA in the evaluation of low levels of factor
Spiked plasma with increasing concentrations of FVIII in the very low VIII and factor IX. However, other sets of parameters can be used for
values impacted clotting time and maximum acceleration (Min2 in the the expression of CWA modifications, as shown in the study led by
study) in a dose-­dependent manner. The Min2 parameter appeared to Milos.13 They expressed, with a parameter called Delta, the steepness
be very well correlated (r = 0.922) with FVIII clotting assay (FVIII:C) of the aPTT curve during the coagulation phase. They then showed
and furthermore had the advantage of being more sensitive than that Delta is significantly different between severe and nonsevere HA
FVIII:C in the very low values (<1.0 IU/dL with FVIII activity clotting patients and that this parameter could be used to discriminate be-
assay). Matsumoto et al extended these assumptions with the testing tween those 2 groups of patients with a remarkable sensitivity and
of aPTT CWA on FVIII and factor IX (FIX)-­deficient plasmas spiked specificity (ROC analysis: area under the curve = 0.978, P < .001; sen-
with their respective factor.2 They showed a good predictable dose-­ sitivity = 97.3%; specificity = 93.2%). Moreover, they showed that the
response of CWA parameters (clotting time, maximum velocity and occurrence of severe clinical events (age of first bleed, number of an-
maximum acceleration) with low amounts of FVIII and FIX. In addition, nual joint bleeds and number of joints with haemophiliac arthroplasty)
they showed that CWA correlated well with TGT across the whole is better correlated with the Delta parameter than with FVIII activity
range of concentrations for FVIII (0-­100 IU/dL) and in the moderate assays (either one-­stage clotting assay or chromogenic assay). This
to high concentrations (1.0-­100 IU/dL) for FIX. According to their ob- confirms the potential of aPTT CWA in the prediction of bleeding phe-
servations, CWA seems to be very accurate, even more than FVIII:C notypes regardless of CWA parameters used to express the results.
clotting assay, for detecting low values (<1.0 IU/dL) of FVIII. Another concept of CWA has been made by Tokunaga et al as
The concept that CWA could be a valuable tool for severe haemo- they observed different profiles in the second-­derivative curves of the
philiac patients has emerged following the repeated observation that aPTT assay: normal pattern, a shoulder type curve and a biphasic type
some patients categorized as severe haemophiliacs, based on FVIII:C curve with a double peak. The presence of the double peak was asso-
assay, exhibit only a moderate bleeding phenotype. Similarly, patients ciated with severity of intrinsic pathway factor deficiency. In addition,
with moderate HA according to the FVIII:C assay can present with in acquired von Willebrand disease patients, the biphasic pattern is
a severe bleeding phenotype characterized by frequent episodes of normalized following DDAVP infusions which increase von Willebrand
spontaneous bleeding.9,10 CWA, through its good correlation with the factor levels.14
FVIII:C assay and its increased sensitivity in the lower values, appears Siegemund et al made a new approach to CWA in 2014. On the
to be a useful prognosis assay for the segregation of patients accord- assumption, as in TGT, that the derivative of a substrate concentration
ing to their bleeding phenotype. The clinical relevance of CWA has curve over time represents the activity of the substrate’s enzyme, they
been assessed by Nair et al in a cohort of severe, moderate and mild considered that the first derivative of the aPTT assay curve represents
HA patients.11 They demonstrated that the median value of maximum thrombin activity, the second derivative represents prothrombinase
acceleration (Max2) could discriminate firstly haemophiliac patients activity, and finally, the third derivative represents tenase activity.15
from healthy patients, and secondly between moderate HA and se- They correlated the maximum absolute value of the first derivative
vere HA. However, as did Shima,1 they showed that Max2 presents with the activity of thrombin calculated with a Michaelis-­Menten
more variability in regard to FVIII:C in severe HA subject samples (ie mathematical model and found an excellent correlation (r2 = 0.95). In
with FVIII:C<1.0 IU/dL) than in spiked samples for the same range of addition, this study is the first to present the third derivative as the
FVIII:C level. This could suggest that FVIII is not the only determinant best parameter correlated with low concentrations of FVIII.
for clot formation acceleration in clinical samples and that factors In view of the interest of these parameters in the evaluation of
other than FVIII could be implicated in the expression of bleeding ten- severe haemophilia A and B, the implementation into clinical practice
dency in HA patients. Based on this assumption, and the fact that it is as an aid for determining the bleeding risk should be considered. It
a global coagulation test, CWA reflects the balance between different could be even more valuable for laboratories which do not have ca-
procoagulant and anticoagulant factors present in the plasma sample pabilities for measuring FVIII:C or FIX:C. However, aforementioned
more effectively, and thus predicts the bleeding tendency for haemo- studies were conducted on small series of patients only. Conducting
philiac patients better than the determination of FVIII:C alone. This larger and prospective studies including reliable collection of clinical
564       SEVENET and DEPASSE

data should be considered to confirm the potential of clot waveform different alloantibodies against different epitopes of FVIII. This helped
in haemophilia. to understand the variability of responses to FVIII replacement ther-
apy in patients with HA with inhibitors.21
It can thus be seen that CWA is a very promising tool for the eval-
3.2 | Replacement therapy management
uation of the effects of haemostatic drugs, either for research and de-
In the course of haemophiliac patient management, development of velopment purposes by pharmaceutical companies or in clinical use for
antibodies targeting FVIII or FIX is the most important complication the follow-­up of patients’ haemostasis recovery.
in replacement therapies. Haemophiliac patients with high responding
inhibitors are nowadays treated for bleeding prevention with bypass-
3.3 | Disseminated intravascular coagulation and sepsis
ing therapies, usually human recombinant activated factor VII (rFVIIa)
or activated prothrombin complex concentrates (aPCC). Optimization Disseminated intravascular coagulation is an acquired disorder of the
of bypassing therapies in these particular patients remains challeng- coagulation and fibrinolytic system, secondary to defined groups of
ing. Neither practical guidelines for optimal therapy nor haemostasis severe diseases such as sepsis or trauma.22
monitoring has been adequately defined. Furthermore, the optimal It was observed by Downey et al in 1997 that a patient in which
dosage to obtain sufficient efficacy with the lowest risk of throm- DIC has been diagnosed could present an abnormal and characteristi-
botic events varies across patients. At present, standard one-­stage cally distinct transmittance waveform appearance on the aPTT assay.23
clotting factor assays do not always adequately reflect haemostasis They noticed that the curve can show an immediate and progressive
in the presence of an inhibitor, and thus, are not relevant enough for decrease in light transmittance after initiation of the aPTT assay, that is
efficient monitoring. Assays evaluating global clotting function, such in the precoagulation phase, while the normal profile is characterized
as automated thromboelastometry, TGT and CWA, may be useful for by a 100% light transmittance prior to the clot formation. This modi-
evaluating and monitoring haemophiliac patients with inhibitors.16-18 fied appearance was designated as the BWP (Figure 2). The presence
Following observations on the correlation between CWA and very of BWP was shown to be independent of the aPTT clotting time, aPTT
low values of FVIII:C levels, it was proposed by Shima and colleagues activator reagent, decreased or increased coagulation factors and is
that this technique could provide a very useful method for monitor- not influenced by anticoagulant therapy such as heparin.24,25
ing HA patients with inhibitors. They showed a clear dose-­response In a second study led by the same author, the correlation be-
relationship between the improvement in clot waveform parameters tween CWA and DIC diagnosis was assessed in a prospective co-
and the ex vivo added amount of rFVIIa. Furthermore, they highlighted hort of 747 patients. The study showed that the sensitivity and
the benefits of this assay as it could be run as a routine coagulation specificity of BWP for DIC diagnosis was 97.6% (95% confidence
assay. They also demonstrated that CWA results are independent of interval [CI]: 85.6%-­99.9%) and 98% (95% CI: 96.6%-­98.9%), re-
pre-­analytical variables such as in vitro coagulation activation. spectively.5 Toh and colleagues demonstrated that the addition of
More recently, Haku et al explored the performances of CWA in Ca2+ in the test sample induces the rapid formation of a precipi-
bypassing therapy monitoring, for both rFVIIa and aPCC. They as- tate that contains C-­reactive protein (CRP) and very low-­density
sessed different triggers for coagulation and found that the best re- lipoprotein (VLDL). The calculated concentration of CRP-­VLDL
sults, after in vivo infusion of bypassing therapy, were yielded when complexes is correlated with turbidity changes in the sample,
aPTT is activated with a mix of low doses of tissue factor and ellagic thus confirming the assumption.6 Furthermore, VLDL complexed
acid. This study confirms the promising potential of CWA in bypass- to CRP has been shown to induce a procoagulant effect through
ing therapy monitoring, even with a modified aPTT trigger reagent. enhancement of the prothrombinase complex, by modification of
Furthermore, they were the first to monitor the real-­time clinical ef- the anionic phospholipid surface.26 The same group prospectively
fectiveness of replacement therapy using CWA, in a small number of assessed CWA in 1,187 patients admitted in the ICU over a 24-­
patients. month period and concluded that CWA predicts DIC outcome more
Using clot waveform, Kasuda et al assessed the effectiveness of effectively than D-­dimer, CRP or VLDL measurements alone.6 The
immune tolerance induction in patients with high responding inhib- results also showed that the appearance of BWP preceded, by an
itors. They observed that, in patients with inhibitors, CWA based on average of 18 hours (range 2-­47 hours), the time of DIC diagnosis
the aPTT assay more effectively reflects clinical improvement after (using Japanese Ministry of Health and Welfare—JMHW—criteria
factor concentrates infusion than the FVIII:C assay.19 for DIC diagnosis). Furthermore, the quantitation of the degree of
Clot waveform analysis is also used to assist in the development abnormality could provide an index for the prediction of the clini-
of new haemostatic drugs, by the evaluation of coagulation param- cal response of DIC to therapy.27 In this cohort, the overall death
eters. Hence, Shirathata assessed, in a phase 1 trial, the potential rate was found to be 44% for patients who are positive for BWP
effect on coagulation and thrombin generation of a new bypassing whereas it was only 25% for patients with a normal aPTT wave-
drug in haemophilia, MC710 (KAKETSUKEN; Kumamoto, Japan). By form.28 Another prospective study over 217 consecutive ICU pa-
observing maximum velocity and maximum acceleration, they found tients demonstrated that the presence of the BWP has a sensitivity
that MC710 had a greater bypassing activity than FEIBA® or rFVIIa.20 of 88% and a specificity of 97% for diagnosis of DIC, as compared
Furthermore, CWA was used by Yada et al to evaluate the effects of to a blind diagnosis based on experts’ opinion.29

F I G U R E   2   The biphasic waveform

pattern could be seen in patients with
disseminated intravascular coagulation
(DIC) (waveform in dotted line)

In non-­ICU hospitalized patients, the BWP showed a moderate and that this convenient assay could facilitate a prompt diagnosis, in
sensitivity of 59.2% or 47.9% for the diagnosis of DIC by ISTH or combination with procalcitonin measurement.37 The combination with
JMHW criteria, respectively. However, specificity was high (95.4% for procalcitonin levels was notably assessed by Zarariah et al and has
both scores) and the presence of the BWP was significantly associated demonstrated that, at optimized thresholds, this combination showed
with DIC (P < .0001) with odds ratios of 29.9 and 19.0 for ISTH and enhanced sensitivity of 79% (95% CI: 64%-­90%), specificity of 96%
JMHW criteria, respectively.30 (95% CI: 93%-­98%) and negative predictive value of 96% (95% CI:
In addition to DIC condition, BWP has been examined among 94%-­98%).38
sepsis-­affected patients. Several studies have shown a positive cor- Altogether, the cumulative evidences confirm the considerable ca-
relation between the presence of BWP and the early development of pacity of CWA in predicting DIC or sepsis. However, this parameter
sepsis in patients admitted to ICU.28,31-33 They showed a modest sen- is not actually used in a clinical setting, although the sensitivity has
sitivity, while the specificity is higher. BWP has thus been proposed been described as elevated (up to 98%).28,29 This may be due to the
to evaluate a high-­risk ICU population, principally in sepsis-­affected actual limited number of laboratories providing CWA and to the lake of
patients. Given the elevated negative predictive value (NPV) deter- large prospective confirmatory studies. However, it could be of great
mined in the studies conducted by Toh (NPV: 91.0%) and Downey interest to implement it in different scores, as the test is inexpensive,
(NPV: 90.6%), the presence of BWP in aPTT waveform could be used easy to perform and has shown high performances for both DIC and
as a marker of sepsis and for ruling-­out sepsis in ICU patients when sepsis.39
the result is normal.
In cardiopulmonary bypass surgery population, a publication of
3.4 | Other potential indications
Delannoy et al demonstrated, in a cohort of 32 patients, that the
presence of a BWP discriminated between sepsis and nonseptic sys- Since the introduction of the assay into clinical studies, CWA has been
temic inflammatory response syndrome (SIRS) with a sensitivity of assessed to detect procoagulant or prohaemorrhagic patients’ status
100% and a specificity of 93% (ROC analysis, area under the curve: in a variety of disorders.
0.948 ± 0.039; P < .01).34 This finding reinforces what Chopin et al de- Lupus anticoagulants (LA) were notably studied with CWA by Su et
scribed in a ICU population of 187 consecutive patients: the biphasic al40 Investigation of patients with antiphospholipid antibodies (APA)
aPTT waveform was able to discriminate between patients with severe was performed using both aPTT and PT waveforms. They showed
sepsis and sepsis shock from patients with SRIS and nonsevere sepsis that the slope of the precoagulation phase of PT assay (called slope
with a sensitivity of 90% (95% CI: 82%-­94%) and a NPV of 92% (95% 1) presents an abnormal result (a difference of more than 2 SD from
CI: 87%-­96%).35 a healthy group) in 61.5% of patients with APA. In contrast, 5.1% of
It has also been reported in a non-­ICU population that the pres- patients without APA but treated by warfarin, and no healthy patients
ence of aPTT BWP is related to the poorest clinical outcomes (infec- (untreated by warfarin) present an abnormal slope 1. However, these
tion prevalence, overall mortality).36 Consecutively, enrolled patients results were highly reagent-­dependent as it has only been observed
who showed positive aPTT BWP also had positive microbial culture with the Simplastin® reagent, and no significant decrease in slope 1
in 67% of cases, whereas only 13% of patients without BWP had pos- was observed with other thromboplastin reagents. Concerning aPTT
itive microbial culture.32 CWA has been assessed in meningococcal assay, none of the extracted parameters consistently distinguished pa-
sepsis-­affected children. Results showed that the presence of BWP tients with APA from other groups of patients, even aPTT clotting time
is associated with the poorest outcomes (prolonged hospital stay) due to its weak specificity. In other studies, abnormal aPTT waveform
566       SEVENET and DEPASSE

has been observed in 25% of patients with APA,41 but the authors did activity of all factors implicated in the clot formation process. In this
not determine any aPTT waveform parameter that would allow the respect, CWA can be considered as a global haemostasis assay. Other
segregation of APA-­positive patients from DIC patients, as the curve global haemostasis assays are thrombin generation and thromboe-
profiles are too similar.23 A recent study observed that absolute Min1 lastometry. They differ greatly from CWA in their operating princi-
value of aPTT assay is decreased in patients with both positive LA and ples. In TEG®/ROTEM®, viscoelastic properties of the fibrin clot are
clinical antiphospholipid syndrome (2.2 ± 0.5; mean value ± SD), and measured whereas thrombin formation rate is detected by mean of a
in patients with LA but without antiphospholipid syndrome (1.6 ± 0.2), fluorescent marker in TGT.49,50
compared to normal patients (3.1 ± 0.1). Furthermore, Min1 for pa- The TEG®/ROTEM® devices evaluate the kinetics and strength of
tients with mild HA or acquired HA is significantly decreased (1.3 ± 0.2 the fibrin clot over the time and can evaluate both prothrombotic and
and 0.9 ± 0.5, respectively) compared to the aforementioned groups. prohaemorrhagic conditions. The most used clinical application is the
However, even if differences are significant, the distribution of Min1 management of bleeding during surgery or invasive procedures. This
values overlaps, and single CWA testing could not distinguish between approach is mentioned in different guidelines for tailoring transfusion
patients in the different groups. An atypical pattern for Min1 and Min2 strategy during surgery.51,52 In addition, the point-­of-­care format and
curves has also been observed in patients with LA, but none in nor- the ease-­of-­use of these devices allow its implementation in the op-
mal patients, using silica-­based aPTT reagent.43 This observation is erating room. Further applications are however possible as numerous
however highly reagent-­dependent as ellagic acid reagent produces a trigger reagents could be used (extrinsic pathway assessment, hepa-
similar atypical pattern in normal patients and in LA-­positive patients. rin levels, fibrinogen measurement and resistance to lysis). Thrombin
Ultimately, many different observations have been made in patients generation could also assess both thrombotic states and bleeding ten-
with LA, and further studies are required to confirm these observa- dencies. Several applications are now entering the scope of clinical
tions and add proof of the interest of CWA in LA laboratory diagnosis. practice. The management of haemophiliac patients has been actively
Most of the time, aPTT-­based CWA assay is used. However, CWA assessed with thrombin generation. Some studies have been per-
could also be valuable with a PT assay. Matsumoto and colleagues formed jointly with CWA for this purpose.1,2,18,42 The 2 approaches
used PT-­derived CWA in a study intended to show the prediction po- are very sensitive in low FVIII:C values in HA patients. Nevertheless,
tential of the bleeding phenotype in patients with acquired anti-­factor both still require clinical validation in larger clinical trials.
V inhibitors.44 They found discriminating differences in the clot time, Clot waveform analysis provides similar information to TGT as it
maximum velocity (Min1) and maximum acceleration (Min2) between correlates with the rate and velocity of thrombin formation, and re-
patients with life-­threatening bleeding and asymptomatic patients. flects the whole process of thrombin generation over time. This con-
CWA, in conjunction with TGT, was notably used to understand the trasts with routine coagulation assays in which clotting time only
mechanisms underlying the variations in the phenotypes of patients reflects coagulation initiation. In severe haemophilia or in the presence
with this condition. of a specific factor inhibitor, the rate of thrombin is decreased53 and
Fibrinogen assessment has also been investigated. The difference this results in statistically significant variations in CWA parameters.2,19
between the plateau phase of precoagulation and postcoagulation for Increased thrombin generation is observed in thrombophilic states;
both aPTT and PT assays correlates well with fibrinogen levels, inde- however, no study correlating CWA parameters and thrombophilic
pendently of clotting times and plasma quality (haemolysis, icterus and states have been published as yet.
lipaemia). Despite the positive results, this method of fibrin measure- Clot waveform analysis is a global assay which presents many ad-
ment is not innovative and will not replace present conventional fibrin- vantages. Based on widely used and very simple assays such as PT and
ogen assays such as the Clauss method, as performances (precision aPTT, it appears to be easily carried out on widely available automated
and accuracy) are clearly superior. Nonetheless, these studies show analysers with optical detection systems, using routine reagents, and
the direct correlation between fibrinogen concentration and the im- is not associated with additional costs for laboratories. In contrast to
pact on CWA parameters.45,46 CWA, thromboelastometric assays and TGT require specific equip-
Clot waveform analysis has been used by Shima’s team to demon- ment. Furthermore, CWA could be run as a regular routine assay on
strate the increased procoagulant state in chronic spontaneous ur- the same sample as PT and aPTT (0.109 or 0.105 mol/L trisodium ci-
ticaria47 and for the exploration, in association with other global trate blood collection tubes), and with a fast turnaround time com-
coagulation assays, of the paradoxical procoagulant effect induced by pared to other global assays.
plasmin.48 As clotting time in aPTT assay occurs when only 5% of thrombin
is generated, CWA enables the exploration of the coagulation process
with more information than the single clotting time. Throughout stud-
4 |  DISCUSSION ies on DIC, it has been shown that the presence of an abnormal pat-
tern associated with DIC is not influenced by pre-­analytical conditions
CWA is based on the recording of an optical signal over time and nor by the presence of an anticoagulant. Furthermore, clotting time
is intended to reflect the whole process of clot formation. Whether has no impact on the presence of the biphasic pattern.
initiated by a contact activator (silica, kaolin, polyphenols, ellagic However, CWA does have some limitations. As in PT and aPTT
acid) or by tissue factor, the waveform pattern reflects the combined assays, the use of whole blood or platelet-­rich plasma is not possible.

Using an optical measurement method, interferences colouring plasma studies are limited to small cohorts, extensive prospective clinical tri-
such as haemolysis, lipaemia and icterus can impact CWA results. als are mandatory to make the leap to clinical application.
These interferences depend on the wavelength assigned for optical
detection. At present, no study has been run to assess the minimal
levels of these interferences to perform CWA in optimal conditions REFERENCES
nor the choice of optimal wavelength for clot detection. 1. Shima M, Matsumoto T, Fukuda K, et al. The utility of activated partial
Many questions remain open on the potential clinical use of CWA thromboplastin time (aPTT) clot waveform analysis in the investiga-
in daily practice. Scientific subcommittees of ISTH proposed standard- tion of hemophilia A patients with very low levels of factor VIII activity
(FVIII:C). Thromb Haemost. 2002;87:436‐441.
ization using a sensitive optical detection system, colourless and non-
2. Matsumoto T, Shima M, Takeyama M, et al. The measurement of low
opaque reagent and an adjusted normal clotting time for aPTT assay.8 levels of factor VIII or factor IX in hemophilia A and hemophilia B
However, many experimental factors such as aPTT reagents, types of plasma by clot waveform analysis and thrombin generation assay. J
activator and phospholipids preparations could impact CWA results. Thromb Haemost. 2006;4:377‐384.
3. Shima M, Matsumoto T, Ogiwara K. New assays for monitoring hae-
As yet no study nor correlation has been published on the variability
mophilia treatment. Haemoph Off J World Fed Hemoph. 2008;14(Suppl
between reagents or analysers on CWA results. The choice of wave- 3):83‐92.
length for detection would seem to be crucial to the sensitivity of the 4. Haku J, Nogami K, Matsumoto T, Ogiwara K, Shima M. Optimal moni-
assay, but it is not dealt with in the SSC communication. Furthermore, toring of bypass therapy in hemophilia A patients with inhibitors by the
use of clot waveform analysis. J Thromb Haemost. 2014;12:355‐362.
different curve smoothing algorithms or curve evaluation modes exist
5. Downey C, Kazmi R, Toh CH. Early identification and prognostic im-
and may be a source of variability between analyser results.54 For it plications in disseminated intravascular coagulation through transmit-
to become a clinically validated tool, further studies and calibration tance waveform analysis. Thromb Haemost. 1998;80:65‐69.
methodologies should be developed. However, in contrast with other 6. Toh CH, Samis J, Downey C, et al. Biphasic transmittance waveform
global assays, CWA appears to be the easiest assay to standardize be- in the APTT coagulation assay is due to the formation of a Ca(++)-­
dependent complex of C-­reactive protein with very-­low-­density lipo-
cause of the easy-to-perform protocol, and convenient and very well-­
protein and is a novel marker of impending disseminated intravascular
known assays involved. coagulation. Blood. 2002;100:2522‐2529.
Despite the accumulated data suggesting benefits in various clin- 7. Braun PJ, Givens TB, Stead AG, et al. Properties of optical data from
ical situations and the many attempts to democratize it, CWA is still activated partial thromboplastin time and prothrombin time assays.
Thromb Haemost. 1997;78:1079‐1087.
only investigated by a small number of research teams and seems to
8. Shima M, Thachil J, Nair SC, Srivastava A. Towards standardization of
be under-­recognized by physicians and pathologists. The number of clot waveform analysis and recommendations for its clinical applica-
original articles present in the literature is relatively low. Between tions. J Thromb Haemost. 2013;11:1417‐1420.
1997 and 2016, only 56 original articles and reviews are referenced 9. Van den Berg HM, De Groot PHG, Fischer K. Phenotypic heterogene-
ity in severe hemophilia. J Thromb Haemost. 2007;5:151‐156.
in MEDLINE/PubMed, compared to more than 6000 and more than
10. Nogami K, Shima M. Phenotypic heterogeneity of hemostasis in se-
4000 when using keywords “thrombin generation” and “thrombelas- vere hemophilia. Semin Thromb Hemost. 2015;41:826‐831.
tography,” respectively. 11. Nair SC, Dargaud Y, Chitlur M, Srivastava A. Tests of global hae-
Shown to be a very promising tool, CWA should be extensively mostasis and their applications in bleeding disorders. Haemophilia.
2010;16(Suppl 5):85‐92.
studied in many different clinical conditions, especially as more and
12. Yada K, Nogami K, Wakabayashi H, Fay PJ, Shima M. The mild phe-
more coagulation analysers provide optical detection system and notype in severe hemophilia A with Arg1781His mutation is associ-
possess integrated software for waveform analysis. Furthers physio- ated with enhanced binding affinity of factor VIII for factor X. Thromb
pathological conditions, such as hypercoagulable states, inflammation, Haemost. 2013;109:1007‐1015.
13. Milos M, Coen Herak D, Zupancic-Salek S, Zadro R. New quantitative
anticoagulant monitoring and risk prediction, have to be extensively
aPTT waveform analysis and its application in laboratory management
evaluated, to integrate this rapid, inexpensive and already available of haemophilia A patients. Haemophilia. 2014;20:898‐904.
tool into clinical practice. 14. Tokunaga N, Inoue C, Sakata T, et  al. Usefulness of the second-­
derivative curve of activated partial thromboplastin time on the ACL-­
TOP coagulation analyzer for detecting factor deficiencies. Blood
Coagul Fibrinolysis. 2016;27:474‐476.
5 | CONCLUSION 15. Siegemund T, Scholz U, Schobess R, Siegemund A. Clot waveform anal-
ysis in patients with haemophilia A. Hämostaseologie. 2014;34(Suppl
It is now generally accepted that CWA has huge potential and could 1):S48‐S52.
provide much more information on the haemostasis system than clot- 16. Young G, Sørensen B, Dargaud Y, Negrier C, Brummel-Ziedins K, Key
NS. Thrombin generation and whole blood viscoelastic assays in the
ting time alone. The most widely used assays are aPTT, and to a lesser
management of hemophilia: current state of art and future perspec-
extent PT or combined tissue factor plus contact phase activator tives. Blood. 2013;121:1944‐1950.
triggered assays. Efforts have been made to standardize the assay; 17. Salvagno GL, Berntorp E. Thrombin generation testing for monitoring
however, further recommendations are still needed for an extensive hemophilia treatment: a clinical perspective. Semin Thromb Hemost.
clinical application. At present, HA management and DIC diagnosis
18. Shima M. Understanding the hemostatic effects of recombinant factor
and prognosis are the 2 main clinical fields in which CWA is designed VIIa by clot wave form analysis. Semin Hematol. 2004;41(1 Suppl 1):
to deliver improvements to patient management. Nevertheless, as 125‐131.
568       SEVENET and DEPASSE

19. Kasuda S, Tanaka I, Shima M, et  al. Effectiveness of factor VIII in- analysis in pediatric meningococcal sepsis. Pediatr Crit Care Med.
fusions in haemophilia A patients with high responding inhibitors. 2011;12:e322‐e329.
Haemophilia. 2004;10:341‐346. 38. Zakariah AN, Cozzi SM, Van Nuffelen M, Clausi CM, Pradier O,
20. Shirahata A, Fukutake K, Mimaya J, et  al. Results of clot waveform Vincent J-L. Combination of biphasic transmittance waveform with
analysis and thrombin generation test for a plasma-­derived factor VIIa blood procalcitonin levels for diagnosis of sepsis in acutely ill patients.
and X mixture (MC710) in haemophilia patients with inhibitors-­phase Crit Care Med. 2008;36:1507‐1512.
I trial: 2nd report. Haemophilia. 2013;19:330‐337. 39. Ten Cate H. The biphasic waveform in plasma: identifying the sepsis-­
21. Yada K, Nogami K, Shima M. Different factor VIII neutralizing effects coagulation crossroad? J Thromb Haemost. 2004;2:1533‐1534.
on anti-­factor VIII inhibitor antibodies associated with epitope speci- 40. Su Z, Braun PJ, Klemp KF, Baker KR, Thames EH, Ortel TL. Abnormal
ficity and von Willebrand factor. Br J Haematol. 2013;163:104‐111. optical waveform profiles in coagulation assays from patients with an-
22. Levi M, Ten Cate H. Disseminated intravascular coagulation. N Engl J tiphospholipid antibodies. Blood Coagul Fibrinolysis. 2002;13:7‐17.
Med. 1999;341:586‐592. 41. Luddington R, Peters J, Baker P. Optical waveform analysis is a very
23. Downey C, Kazmi R, Toh CH. Novel and diagnostically applicable in- sensitive indicator of disseminated intravascular coagulation but not
formation from optical waveform analysis of blood coagulation in dis- 100% specific. Br J Haematol. 1997;99:469‐470.
seminated intravascular coagulation. Br J Haematol. 1997;98:68‐73. 42. Matsumoto T, Nogami K, Shima M. A combined approach using
24. Toh CH. APTT revisited: detecting dysfunction in the hemostatic sys- global coagulation assays quickly differentiates coagulation disorders
tem through waveform analysis. Thromb Haemost. 1999;82:684‐687. with prolonged aPTT and low levels of FVIII activity. Int J Hematol.
25. Toh CH, Giles AR. Waveform analysis of clotting test optical profiles 2016;105:174‐183.
in the diagnosis and management of disseminated intravascular coag- 43. Solano C, Zerafa P, Bird R. A study of atypical APTT derivative
ulation (DIC). Clin Lab Haematol. 2002;24:321‐327. curves on the ACL TOP coagulation analyser. Int J Lab Hematol.
26. Dennis MW, Downey C, Brufatto N, Nesheim ME, Stevenson K, Toh 2011;33:67‐78.
CH. Prothrombinase enhancement through quantitative and qualita- 44. Matsumoto T, Nogami K, Shima M. Coagulation function and mecha-
tive changes affecting very low density lipoprotein in complex with nisms in various clinical phenotypes of patients with acquired factor V
C-­reactive protein. Thromb Haemost. 2004;91:522‐530. inhibitors. J Thromb Haemost. 2014;12:1503‐1512.
27. Samis JA, Stewart KA, Toh CH, Day A, Downey C, Nesheim ME. Temporal 45. Rossi E, Mondonico P, Lombardi A, Preda L. Method for the determi-
changes in factors associated with neutrophil elastase and coagulation nation of functional (clottable) fibrinogen by the new family of ACL
in intensive care patients with a biphasic waveform and disseminated coagulometers. Thromb Res. 1988;52:453‐468.
intravascular coagulation. J Thromb Haemost. 2004;2:1535‐1544. 46. Tan V, Doyle CJ, Budzynski AZ. Comparison of the kinetic fibrinogen
28. Toh CH, Ticknor LO, Downey C, Giles AR, Paton RC, Wenstone R. assay with the von Clauss method and the clot recovery method in
Early identification of sepsis and mortality risks through simple, rapid plasma of patients with conditions affecting fibrinogen coagulability.
clot-­waveform analysis. Implications of lipoprotein-­complexed C re- Am J Clin Pathol. 1995;104:455‐462.
active protein formation. Intensive Care Med. 2003;29:55‐61. 47. Takeda T, Sakurai Y, Takahagi S, et al. Increase of coagulation potential
29. Bakhtiari K, Meijers JCM, de Jonge E, Levi M. Prospective validation in chronic spontaneous urticaria. Allergy. 2011;66:428‐433.
of the International Society of Thrombosis and Haemostasis scoring 48. Ogiwara K, Nogami K, Nishiya K, Shima M. Plasmin-­induced proco-
system for disseminated intravascular coagulation. Crit Care Med. agulant effects in the blood coagulation: a crucial role of coagulation
2004;32:2416‐2421. factors V and VIII. Blood Coagul Fibrinolysis. 2010;21:568‐576.
30. Matsumoto T, Wada H, Nishioka Y, et al. Frequency of abnormal bi- 49. Hemker HC, Al Dieri R, De Smedt E, Béguin S. Thrombin genera-
phasic aPTT clot waveforms in patients with underlying disorders tion, a function test of the haemostatic-­thrombotic system. Thromb
associated with disseminated intravascular coagulation. Clin Appl Haemost. 2006;96:553‐561.
Thromb Hemost. 2006;12:185‐192. 50. Ganter MT, Hofer CK. Coagulation monitoring: current techniques
31. Dempfle C-EH, Lorenz S, Smolinski M, et al. Utility of activated par- and clinical use of viscoelastic point-­of-­care coagulation devices.
tial thromboplastin time waveform analysis for identification of sepsis Anesth Analg. 2008;106:1366‐1375.
and overt disseminated intravascular coagulation in patients admitted 51. Kozek-Langenecker SA, Afshari A, Albaladejo P, et al. Management of
to a surgical intensive care unit. Crit Care Med. 2004;32:520‐524. severe perioperative bleeding: guidelines from the European Society
32. Smith EY, Charles LA, Van Cott EM. Biphasic activated partial throm- of Anaesthesiology. Eur J Anaesthesiol. 2013;30:270‐382.
boplastin time waveform and adverse events in non-­intensive care 52. American Society of Anesthesiologists Task Force on Perioperative
unit patients. Am J Clin Pathol. 2004;121:138‐141. Blood Management. Practice guidelines for perioperative blood
33. Hussain N, Hodson D, Marcus R, Baglin T, Luddington R. The biphasic management: an updated report by the American Society of
transmittance waveform: an early marker of sepsis in patients with Anesthesiologists Task Force on Perioperative Blood Management*.
neutropenia. Thromb Haemost. 2008;100:146‐148. Anesthesiology. 2015;122:241‐275.
34. Delannoy B, Guye M-L, Slaiman DH, Lehot J-J, Cannesson M. Effect 53. Kattula S, Byrnes JR, Wolberg AS. Fibrinogen and Fibrin in Hemostasis
of cardiopulmonary bypass on activated partial thromboplastin time and Thrombosis. Arterioscler Thromb Vasc Biol. 2017;37:e13‐e21.
waveform analysis, serum procalcitonin and C-­reactive protein con- 54. Milos M, Herak DC, Zadro R. Discrepancies between APTT results de-
centrations. Crit Care. 2009;13:R180. termined with different evaluation modes on automated coagulation
35. Chopin N, Floccard B, Sobas F, et al. Activated partial thromboplastin analyzers. Int J Lab Hematol. 2010;32:33‐39.
time waveform analysis: a new tool to detect infection?*. Crit Care
Med. 2006;34:1654‐1660.
36. Mair G, Dunhill S, Tiplady C. Prognostic implications of a biphasic
How to cite this article: Sevenet PO, Depasse F. Clot
waveform for APTT analysis in a district general hospital. Int J Lab
waveform analysis: Where do we stand in 2017?. Int J Lab
Hematol. 2008;30:467‐472.
37. Paize F, Carrol E, Downey C, et  al. Diagnostic efficacy of acti- Hem. 2017;39:561–568. https://doi.org/10.1111/ijlh.12724
vated partial thromboplastin time waveform and procalcitonin