Int J Lab Hematology - 2015 - Adcock - Pearls and Pitfalls in Factor Inhibitor Assays

International Journal of Laboratory Hematology
The Official journal of the International Society for Laboratory Hematology
REVIEW INTERNAT IONAL JOURNAL OF LABORATO RY HEMATO LOGY
Pearls and pitfalls in factor inhibitor assays

D. M. ADCOCK*, E. J. FAVALORO †
*Colorado Coagulation, S U M M A RY
Laboratory Corporation of
Americaâ Holdings, Englewood, The proper performance and interpretation of factor inhibitor
CO, USA assays is a critical role for the hemostasis laboratory. Both false-
†
Department of Haematology,
Institute of Clinical Pathology positive and false-negative inhibitor assays may be reported, lead-
and Medical Research, ing to serious patient mismanagement. Knowledge and recognition
Pathology West, NSW Health of common causes of both false-positive and negative-results can
Pathology, Westmead Hospital,
aid in the identification of these potential pitfalls. Safeguards to
Sydney, NSW, Australia
reporting accurate factor inhibitor assays include initial character-
Correspondence: ization of the sample, using the Nijmegen modification, properly
Dorothy M. Adcock, Colorado performing and interpreting an incubated mixing test in conjunc-
Coagulation, Laboratory
tion, and performing two dilutions for each dependent dilution in
Corporation of Americaâ Hold-
ings, 8490 Upland Dr., Engle- the factor inhibitor assay.
wood, CO 80112-7116, USA.
Tel.: 720 568 4328;
Fax: 720 568 4324;
E-mail: [email protected]
doi:10.1111/ijlh.12352
Received 2 January 2015;

accepted for publication 10
March 2015
Keywords
Factor inhibitor assay, false
positive, false negative, factor
VIII inhibitor, lupus
anticoagulant, EDTA
tion of coagulation test(s), although not all antibodies

INTRODUCTION
are functionally neutralizing [2]. The specific coagula-
Detection and proper characterization of factor inhibition test(s) prolonged by functional inhibitors depends
tors is essential to patient management yet remains a on the factor(s) inhibited and its location within the
challenge for most diagnostic laboratories [1]. Factor coagulation cascade. Factor inhibitors may develop in
inhibitors are polyclonal immunoglobulins that bind patients with congenital factor deficiency in response
and neutralize the activity of a coagulation factor to factor replacement therapy; those against factor VIII
resulting in deficiency of factor activity and prolonga- (FVIII) and factor IX (FIX) comprise the vast majority
52 © 2015 John Wiley & Sons Ltd, Int. Jnl. Lab. Hem. 2015, 37 (Suppl. 1), 52–60
1751553x, 2015, S1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/ijlh.12352 by Cochrane Saudi Arabia, Wiley Online Library on [16/06/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
D. M. ADCOCK AND E. J. FAVALORO | FACTOR INHIBITOR ASSAYS 53
of these alloantibodies. Early detection is crucial to 37 °C for 60–120 min, are indicated in the evaluation
optimized patient care; thus, surveillance for inhibitor of a prolonged activated partial thromboplastin time
development in patients under factor replacement (APTT) when a FVIII inhibitor is in question, such as
treatment is regularly undertaken [3]. Factor inhibi- in a patient with hemophilia A (i.e. congenital FVIII
tors may also develop as an autoimmune phenome- deficiency) or a patient with a recent bleeding diathe-
non in individuals without congenital factor sis. See Figure 1 for proper performance of an incu-
deficiency [4, 5]. The most frequent autologous inhib- bated mixing test [15]. A time- and temperature-
itor is directed against FVIII, and this may lead to a dependent inhibitor is present when the results of the
spontaneous, life-threatening bleeding diathesis [6]. patient/normal pool plasma (NPP) mixture are longer
Development of inhibitors in either circumstance than the control by a given number of seconds or per-
necessitates specific therapy in a timely fashion. centage, such as, 20%. Prolongation with incubation
Patient management can be clinically difficult, as they is characteristic of FVIII inhibitors and does not occur
may not respond to conventional therapies. The with other specific or nonspecific coagulation factor
hemostasis laboratory plays a key role in accurate inhibitors, with the possible exception of FV inhibitors
identification and quantitation of factor inhibitors. [16]. A false-positive incubated mixing test may occur
The laboratory assay used to measure factor inhibi- when EDTA plasma is mistakenly used in place of
tors was standardized at a 1975 meeting in Bethesda, sodium citrate plasma [12]. While it is reported that
Maryland, and hence is called a Bethesda Assay (BA) up to 15% of lupus anticoagulants (LA) are time- and
[7]. Modifications of the original assay, referred to as
the Nijmegen–Bethesda assay (NBA), have been intro- Incubated mixing test
duced to enhance standardization and decrease the
Incubate each at
incidence of false-positive results [8]. However, signif- 37° C for 60 to
icant numbers of false-positive (up to 32%) and false- 120 min, then
50% 50% mix paent with
negative (up to 5%) results are reported, evidenced NPP
Paent NPP
by various external proficiency surveys, with little Paent + NPP
improvement demonstrated over the past decade [1,

9–14]. This most likely reflects the variety of inhibitor Compare incubated
mix to control
assay methods used by different laboratories, lack of
incorporation of the Nijmegen modification, poor
Separately incubated Paent + NPP
appreciation of potential assay interferences and their paent + NPP mix Incubated together (Incubated mix)
impact, as well as incomplete characterization of sam- (Control)
ples prior to inhibitor testing. False identification of

the presence or absence of a factor inhibitor may have Figure 1. In the proper performance of an incubated
tremendous impact on patient care. This article begins mixing test, three plasma aliquots are prepared. The
with a review of the incubated mixing assay as its first aliquot is patient sample, the second aliquot is
NPP, and the third is patient sample mixed with an
proper performance and interpretation is integral to
equal volume of NPP. All three aliquots are held at
the accurate laboratory evaluation of factor inhibitors. 37 °C for 60–120 min in a water bath. Following
This is followed by a review of the BA methodology incubation, the separately incubated patient sample
including the Nijmegen modification. We then focus and NPP are combined in equal volume (1 : 1) and
on the potential causes of false-positive and false-neg- an activated partial thromboplastin time performed
ative factor inhibitor assays, concluding with recom- on this mixture (which represents the control), as
well as the patient/NPP mixture that was incubated
mendations on how to avoid such misclassifications.
(which represents the test). The time in seconds of
the control is compared to the incubated patient/NPP
mixture. A time- and temperature-dependent
I N C U BAT E D N O R M A L P L A S M A M I X I N G T E S T
inhibitor is present when the results of the patient/
These are often performed and interpreted incorrectly NPP mixture are longer than the control by a given
and under inappropriate clinical conditions. Incubated number of seconds or percentage, such as, 20%.
NPP, normal pool plasma.
mixing studies, in which a plasma sample is held at
© 2015 John Wiley & Sons Ltd, Int. Jnl. Lab. Hem. 2015, 37 (Suppl. 1), 52–60
54 D. M. ADCOCK AND E. J. FAVALORO | FACTOR INHIBITOR ASSAYS
temperature-dependent, this likely represents an factor activity or calculated based on the following
artifact due to change in pH with incubation, rather formula: BT = (2 log %RA)0.30; being certain to
than a characteristic of the inhibitor [17]. correct for dilutions if needed, where BT stands for
Bethesda titer and RA (residual activity) [21]. A BT
should be calculated only when the RA falls between
BA M E T H O D O L O G Y I N C L U D I N G T H E
25% and 75%. A RA of 75% correlates to a BT of
N I J M E G E N M O D I F I C AT I O N
0.4 BU, and therefore, this is the lowest possible BT
Adaptations of BA methodology can be used to identify that can be reported using this method. When the RA is
not only FVIII inhibitors, but also inhibitors to other >75%, the BT is reported as negative. If the RA falls
coagulation factors including FXI, FX, FIX, FVII, FV, below 25%, additional dilutions of the patient plasma
and FII, as well as FXIII, protein S and von Willebrand should be tested. When multiple residual activities fall
factor activity [18]. In the original BA assay procedure, between 25% and 75%, the least dilute sample closest
a mixture of patient plasma and NPP incubated at to 50% is used to determine the BT.
37 °C for 2 h is compared to a control plasma (created The BA is often performed using multiple dilutions
by combining NPP with an equal volume of buffer) that of patient plasma in buffer so that a BT can be deter-
is subject to the same assay conditions [7]. Samples mined. Patient plasma is typically tested neat, at 1 : 2
must remain capped whenever feasible to help main- and 1 : 5 dilutions, although additional dilutions may
tain pH [19]. Factor activity of the patient mix and con- be required with high-titer inhibitors. These initial
trol plasma is measured following the incubation. For patient dilutions are referred to as ‘dependent dilu-
detection of inhibitors other than FVIII inhibitors, a tions’. When the factor activity is performed on these
10 min- to 1-h incubation at 37 °C is generally consid- dependent dilutions following incubation, the analyzer
ered sufficient [20]. The residual percent factor activity should be programmed to perform the activity assay at
is determined and expressed as a ratio of the patient 1 : 10 and 1 : 20 dilutions and these results evaluated
mix result over the control (see Figure 2). A Bethesda for nonparallelism (Table 1). This step will aid in the
unit (BU) is defined as the amount of inhibitor that will detection of nonspecific inhibitor effect, which could
inactivate half of the factor present (i.e. results in 50% potentially result in a false-positive BT (Tables 2 and
residual factor activity) in an equal mixture of patient 3). Some nonspecific inhibitors that cannot be distin-
and NPP following incubation at 37 °C for 2 h [7]. guished in a factor activity assay may be detected in the
If an inhibitor is present, inhibitor titer can be deter- dilutions of the BA. This enhanced sensitivity to non-
mined using a graph relating inhibitor units to residual specific inhibitors in the BA over a factor activity assay
may be due to the added NPP and 37 °C incubation. As
an example, when a strong LA sample without a FVIII
Bethesda assay methodology inhibitor is tested in a FVIII activity assay, it may result
Paent mix Control mix
in no measurable FVIII activity at all dilutions tested. A
Make diluons of paent plasma
(“dependent diluons”) 1:1 Mix buffer and NPP FVIII BA performed on this same sample would appear
1:1 Mix diluons of paent and NPP
positive; however, the 1 : 10 and 1 : 20 dilutions of
Incubate each mix 1–2 h 37 °C water bath dependent dilutions typically show nonparallelism.
Measure factor acvity at 1:10 and 1:20 diluons
The International Society on Thrombosis and
% Residual acvity (RA) = (Paent mix ÷ Control mix) X 100
Hemostasis (ISTH) advocates that all laboratories use
Bethesda Unit (BU) = 2–log %RA ÷ 0.30; correct for diluons if needed the NBA over than traditional BA [22]. The Nijmegen
Nijmegen modificaon uses buffered normal plasma in the paent mix modification introduces the following two variations:
and FVIIIDP* or 4% albumin rather than buffer in the conrol mix
(i) the use of FVIII-deficient plasma (or 4% BSA)
FVIIIDP: Factor VIII Depleted Plasma
rather than buffer in the control sample and (ii) buf-
fering of NPP as used in both the patient and control
Figure 2. Methodology of the Bethesda assay mixtures [8, 23]. A lyophilized NPP can be reconsti-
including the Nijmegen modification. *FVIIIDP, tuted with imidazole buffer to create a buffered NPP.
factor VIII depleted plasma. NPP, normal pool External quality Control of Assays and Tests (ECAT)
plasma.
and Royal College of Pathologists of Australasia
Table 1. Example of a Bethesda assay calculation sheet with a high-titer factor VIII (FVIII) inhibitor
FVIII Bethesda assay calculation sheet
Patient tube FVIII

activity
Dependent Control Log residual BU 9 dilution
Patient dilution 1 : 10 1 : 20 Av factor activity %RA activity BU factor
Smith 1 0 0 0 45 0 ncalc ncalc ncalc

20 0 0 0 45 0 ncalc ncalc ncalc
40 5.2 0 3 45 7 0.82 3.9 156
80 8.1 9.9 9 45 20 1.30 2.3 186
160 11.7 12.7 12 45 27 1.43 1.9 305
320 16.0 16.3 16 45 36 1.55 1.5 477
640 23.1 25.2 24 45 53 1.73 0.9 580
1280 33.3 32.0 33 45 73 1.87 0.45 573
RA, residual activity; BU, Bethesda unit; ncalc, unable to calculate; av, average.
Example of a factor VIII inhibitor calculation sheet using the following formula: BT = (2 log %RA)/0.30 and correct-
ing for the dependent dilutions. This factor VIII inhibitor has a titer of 580 BU. Italics indicated the value of the titer.
Table 2. Example of a factor IX Bethesda assay calculation sheet with a high-titer factor VIII inhibitor without
performing two dilutions for each dependent dilution resulting in a false-positive factor IX (FIX) inhibitor titer
FIX Bethesda assay calculation sheet
Patient tube Control factor Log residual BU 9 dilution

Patient Dilution FIX activity activity %RA activity BU factor
Jones 1 0 53 0 ncalc ncalc ncalc

2 0 53 0 ncalc ncalc ncalc
5 9.3 53 18 1.24 2.51 12.6
10 15.8 53 30 1.47 1.75 17.5
20 24.4 53 46 1.66 1.12 22.4
40 37.2 53 70 1.85 0.51 20.4
80 42.5 53 80 1.90 0.32 25.5
160 50.4 53 95 1.98 0.07 11.6
ncalc, unable to calculate; RA, residual activity; BU, Bethesda unit.

Example of a factor IX inhibitor calculation sheet using the following formula: BT = (2 log %RA)/0.30 and correct-
ing for the dependent dilutions. This strong factor VIII inhibitor gives a factitious FIX inhibitor titer of 22.4 BU. Italics
indicated the value of the titer.
(RCPA) surveys indicate that only about 30% of par- proteins. These nonspecific inhibitors of coagulation
ticipants use the Nijmegen modification [10, 14]. may interfere with APTT-based factor activity assays
(i.e. FVIII, FIX, FXI, and FXII) causing one or more
spuriously low or undetectable factor activity results
P OT E N T I A L C AU S E S O F FA L S E - P O S I T I V E BA S
[10, 24, 25]. Lupus anticoagulants inhibitor effect in
factor assays can often be diluted with saline or buffer
Lupus anticoagulants
leading to increasing factor activity at the 1 : 10,
Lupus anticoagulants are a heterogeneous population 1 : 20, and 1 : 40 dilutions (nonparallelism) [26]. Par-
of antibodies directed against phospholipid-binding ticularly strong LA, however, may not dilute out
Table 3. Example of a factor IX (FIX) Bethesda assay calculation sheet with a high-titer factor VIII (FVIII) inhibitor
demonstrating nonparallelism in the two dilutions for each dependent dilution suggesting further investigation
before reporting
FIX Bethesda assay calculation sheet
Patient tube FIX

activity
Control factor Log residual BU 9 dilution
Patient Dilution 1 : 10 1 : 20 Av activity %RA activity BU factor
Jones 1 0 0 0 53 ncalc ncalc ncalc ncalc

2 0 0 0 53 ncalc ncalc ncalc ncalc
5 9.3 17.0 * 53 ncalc ncalc ncalc ncalc
*Nonparallelism evident, further investigation needed to rule out nonspecific inhibitor effect; ncalc, unable to calculate;
RA, residual activity; BU, Bethesda unit.
Example of a factor IX inhibitor calculation sheet using the following formula: BT = (2 log %RA)0.30 and correct-
ing for the dependent dilutions. This strong factor VIII inhibitor demonstrates nonparallelism in the 1 : 10 and 1 : 20
dilutions of each FIX activity-dependent dilution indicating a nonspecific inhibitor effect which reflects the FVIII inhibi-
tor neutralizing the FVIII in the FIX-deficient plasma used in the assay.
resulting in one or more intrinsic factor activity assays temperature-dependent APTT prolongation, while LA
to appear low or even undetectable. This interference tend to act as immediate inhibitors. [25]. High-titer
can also produce false-positive BA results [9]. The lab- FVIII inhibitors, however, can be ‘fast acting’ (or high
oratory distinction between an LA and specific intrin- levels capable of swamping the available FVIII) and
sic factor inhibitor can be difficult [10, 25]. In demonstrating inhibition with immediate NPP mix
proficiency surveys conducted by the National Exter- with little if any further prolongation with incubation.
nal Quality Assessment Service (NEQAS), LA positive Further complicating the issue is that FVIII inhibitors
(and FVIII inhibitor negative) samples were incor- may yield false-positive results in some LA assays, spe-
rectly classified as a FVIII inhibitor by 6.5% of partici- cifically the hexagonal phospholipid assay. Dilute Rus-
pants in 1998 and 4.4% in 2007 [13]. Clinically, this sell viper venom (dRVVT)-based LA assays are not
is a critically important distinction and such misinter- interfered by FVIII inhibitors, and thus, a positive LA
pretation can significantly influence patient manage- test result by dRVVT testing can be used with some con-
ment. Patients with LA are potentially at an increased fidence to distinguish LA and FVIII inhibitors. Patients
risk for thrombosis, while factor inhibitors generally with FVIII inhibitors can nevertheless have coincident
lead to a bleeding diathesis. LA [27]. In the distinction between true FVIII inhibitors
One clue to LA interference in the BA is the dem- and spurious positive results due to LA, chromogenic
onstration of nonparallelism in the 1 : 10 and 1 : 20 FVIII activity and ELISA-based FVIII inhibitor assays
dilution results of the dependent dilutions as evi- are extremely useful as neither assay demonstrates LA
denced by increasing activity with increasing dilution. interference. Naturally, additional approaches and clues
Another clue is that multiple intrinsic factor activities can help distinguish LA and specific factor inhibitors,
may appear to be reduced and multiple intrinsic fac- with clinical presentation being one key ingredient –
tors may have positive BT. It is also useful to pay patients with LA are predisposed to thrombosis, while
attention to the results of the incubated mixing study those with specific factor inhibitors would classically
as FVIIII inhibitors typically demonstrate time- and present with a bleeding diathesis.
as only FVIII inhibitors demonstrate time and tempera-

Specific factor inhibitor present but is other than the
ture dependence.
factor inhibitor under investigation
High-titer factor inhibitors may interfere with factor

EDTA plasma
activity assays other than the factor for which the inhib-
itor is directed [28]. For example, a high-titer FVIII Sodium citrate is the sample matrix required for fac-
inhibitor may inhibit the FVIII present in the normal tor inhibitor assays. Samples collected in potassium
plasma used in a FIX activity assay, prolonging the EDTA must result in sample rejection. While it is easy
APTT, leading to a spuriously low FIX activity and posi- to identify the incorrect matrix when samples are
tive FIX BA. High-titer FVIII inhibitors occasionally may submitted as primary tubes, EDTA plasma, serum,
not be able to be diluted out and may result in very low and citrate plasma appear identical in secondary
or even undetectable FIX, FXI, or FXII activity at the aliquots.
1 : 10, 1 : 20, and 1 : 40 dilutions. This may also result EDTA is a very strong calcium chelator resulting in
in false-positive FIX, XI, and/or XII BAs. If the labora- spuriously low FVIII and FV activity results and mod-
tory is requested to perform, for example, only a FIX erate prolongation of both the APTT and PT [19].
BA on this FVIII inhibitor-containing sample, they may Incubated mixing studies using EDTA plasma result in
report out a false-positive FIX inhibitor result (see time- and temperature-dependent prolongation in
Tables 2 and 3). A similar situation may also occur both APTT and PT mixing tests [12]. Furthermore,
when laboratories encounter more unusual factor EDTA plasma may yield false-positive FVIII and FV
inhibitors (e.g. to FV, FX, or FXI). When testing samples BTs, typically in the range of 1–5 BU. In a 2005 RCPA
with these more unusual inhibitors, a laboratory may QAP, 68.3% of 42 laboratories falsely identified a FV
recognize the presence of an inhibitor, but have a and/or FVIII inhibitor in normal EDTA plasma [11].
request to test only for the more common inhibitors, If a FV or FVIII BA is performed in EDTA plasma,
such as FVIII or FIX and therefore the potential to dependent dilutions run at 1 : 10 and 1 : 20 dilutions
report either of these as positive. This has a greater like- do not demonstrate nonparallelism. As EDTA plasma
lihood of occurrence when more common inhibitors are demonstrates prolongation of the APTT, prolongation
tested in a local laboratory, and more esoteric inhibitor with an incubated mixing test, decreased FVIII activity
assays are sent to a reference laboratory. A 2005 RCPA and a positive BA, EDTA plasma can perfectly mimic
QAP survey distributed a strong factor V inhibitor sam- a FVIII inhibitor.
ple (complicated by a coincident LA) to 42 laboratories. Clues to the presence of a false-positive FVIII
Only 63.4% of participants identified the factor V inhib- inhibitor secondary to an otherwise normal EDTA
itor, 4.8% identified a FVIII inhibitor, and 17.1% identi- plasma include a clinically unexpected result; for
fied LA only [11]. If, for example, a specific FVIII example, a positive FVIII inhibitor in a pediatric
inhibitor was subject to a FIX BA, the 1 : 10 and 1 : 20 female patient with no history of bleeding. Another
dilution results of the dependent dilutions typically clue is that EDTA plasma results in prolongation of
demonstrate nonparallelism. Another clue would be both the APTT and PT. FVIII inhibitors prolong the
that multiple intrinsic factors (i.e. FVIII, FIX, FXI, FXII), APTT only, although factor V inhibitors elevate both
if tested, may have positive BTs, with the highest titer the APTT and PT. A chromogenic FVIII activity is fal-
pointing to the true specific factor inhibitor. For exam- sely low when performed with EDTA plasma. An
ple, a high-titer FVIII inhibitor may yield a FVIII BT of ELISA-based FVIII inhibitor assay would be negative
1200 BU, but FIX and FXI BAs performed on the same with otherwise normal EDTA plasma. To confirm the
sample would have titers in the range of 1–20 BU. Eval- sample is EDTA rather than citrated plasma, standard
uation of the results of multiple factor assays may pro- chemistry parameters including potassium, sodium,
vide another clue pointing to the true specific factor and calcium can be performed. EDTA plasma yields
inhibitor. For example, a FVIII inhibitor is likely if the absent calcium, elevated potassium, and normal
FVIII activity is undetectable, but FIX, FXI, and FXII sodium levels, while sodium citrate plasma yields ele-
activities are each low but measurable. Attention to the vated sodium, detectable calcium, and normal potas-
results of the incubated mixing test may also be useful sium levels [29].
Tests reported false identification of an inhibitor by

Anticoagulant therapy
8–17% of participants when testing FVIII-deficient
Novel oral anticoagulants including both direct plasma [10]. A false-positive FVIII inhibitor can be
thrombin and FXa inhibitors are being increasingly reported in a patient with a high-level FV inhibitor, as
used for long-term anticoagulation. These anticoagu- described earlier.
lants, as well as heparin, may cause spuriously
decreased factor activity results and falsely positive
P OT E N T I A L C AU S E S O F FA L S E LY N E G AT I V E
BAs, depending on anticoagulant concentration and
N I J M E G E N – B E T H E S DA T I T E R R E S U LT S
responsiveness of the APTT or PT reagent to these
agents [30–33]. Activated partial thromboplastin time
Recent treatment with factor or by-passing agent
and PT reagents vary significantly in their respon-
siveness to each of these anticoagulant drugs [32– In factor-deficient patients receiving replacement ther-
34]. For this reason, the potential impact of these apy, inhibitor testing is recommended when there is
agents on factor activity and inhibitor assays depends poor clinical response to treatment or if post-therapy
on the specific drug, its plasma concentration, and factor levels are lower than expected. Inhibitor tests
the APTT and/or PT reagent in use. It is inappropri- are most sensitive after the patient’s factor level has
ate to perform a factor inhibitor assay in the returned to baseline for 24 h. Residual factor may
presence of a direct IIa or Xa inhibitor anticoagulant mask or extinguish a low-titer inhibitor. If residual
or heparin that has not been neutralized in the concentrate is present in the patient sample, inactiva-
laboratory. tion of residual factor, for example, by heating the
In the emergency management of a bleeding plasma to 58 °C for 90 min, is recommended. Resid-
patient taking a direct thrombin or Xa inhibitor anti- ual factor antigen following heat inactivation, how-
coagulant, the patient may be mistakenly diagnosed ever, may still interfere with the BA, yielding a lower
as suffering with a factor inhibitor. This is especially BT than if the assay were performed in the absence of
likely if medication history is unavailable or the treat- residual heat-inactivated factor.
ing physician does not understand the effect of these
anticoagulant drugs on coagulation assays. Heparin,
Laboratory error/analytical issue
direct thrombin, and direct Xa inhibitors at significant
concentration may lead to elevation of the APTT with EQA data also report the failure of laboratories to rec-
lack of correction in a 1 : 1 plasma mix [16]. Intrinsic ognize the presence of an inhibitor when one is pres-
factor activities may be falsely low, although nonpar- ent. This is most likely to occur when low-level
allelism can often, but not consistently, be demon- inhibitors of approximately 1.0 BU/mL are present
strated [30, 31]. A BA yields a positive inhibitor titer and if a classic BA is used rather than the Nijmegen
although the 1 : 10 and 1 : 20 dilutions of the depen- modification [1, 14]. Interestingly, differential buffer-
dent dilutions should demonstrate nonparallelism ing used by different laboratories (e.g. in house vs.
[30–32]. commercial buffered normal plasma) will also differ-
Clues to the presence of an anticoagulant drug lie entially affect whether or not laboratories correctly
in the results of screening assays. More information is identify these low-titer inhibitors [1, 18]. Laboratories
available in Assessing Nonvitamin K Antagonist Oral Anti- should be encouraged to evaluate their own assay’s
coagulants (NOACs) in the Laboratory in this issue. low-level sensitivity.
Laboratory error/analytical issue S A F E G UA R D S TO R E P O RT I N G FA L S E -

P O S I T I V E O R FA L S E - N E G AT I V E BA S
False-positive factor inhibitor identification in normal
plasma and/or serum as well as in factor-deficient There are a number of steps laboratories can take to
plasma has been reported and occurs more readily help safeguard against reporting falsely positive or
when a standard BA rather than Nijmegen-modified negative factor inhibitor assays. One is to characterize
assay is used. External quality Control of Assays and all samples by measuring an APTT, PT, and thrombin
clotting time prior to performing inhibitor assays. This Likewise, a sample that demonstrates time- and
is especially useful in reference laboratories that temperature-dependent prolongation warrants mea-
receive secondary aliquot samples often without inclu- suring FVIII activity if not already ordered. Impor-
ded history. Sample characterization will also help in tant caveats include EDTA plasma, novel oral
identifying serum, EDTA plasma, or anticoagulant anticoagulants, and some LA, which can mimic FVIII
therapies, all of which may cause a false-positive inhibitors. A chromogenic FVIII activity does not
inhibitor result. The presence of direct Xa inhibitor show LA interference and can be useful, as can be a
anticoagulants is more difficult to identify than hep- dRVVT to help determine LA.
arin or direct thrombin inhibitor anticoagulants using Implementation of the Nijmegen modification of the
routine clotting assays, as some agents, particularly BA is recommended to avoid reporting false-positive
apixaban, may have little effect on the APTT and PT. and false-negative inhibitor results. In addition, it is
A chromogenic anti-Xa assay can detect the presence useful to measure baseline factor activity prior to the
of direct Xa inhibitor agents. Laboratories without BA to determine whether heat inactivation is needed. If
the capability to run anti-Xa assays may need to rely the baseline activity is too high, a BA is not indicated.
on the patient’s medication history. A final safeguard in performing the BA is to
Proper performance and interpretation of incu- include 1 : 10 and 1 : 20 dilutions in the factor activ-
bated mixing tests and correlation to inhibitor assays ity assay performed on the dependent dilutions. Non-
results is important. An incubated mixing test is parallelism suggests a nonspecific inhibitor and the
indicated when a FVIII inhibitor is under investiga- need to further investigate the sample.
tion. If a sample with a prolonged APTT demon-
strates correction or near correction on addition of
CONFLICT OF INTEREST
normal plasma, but no further prolongation with
incubation, a FVIII inhibitor is likely not present. The authors declare no conflict of interest.
REFERENCES 7. Kasper CK, Aledort LM, Counts BB, Edson 11. Favaloro EJ, Bonar R, Duncan E, Earl G,
JR, Frantantoni J, Green D, Hamptin JW, Low J, Aboud M, Just S, Sioufi J, Street A,
1. Favaloro EJ, Meijer P, Jennings I, Sioufi J, Hilgartner MW, Lazerson J, Levine PH, Marsden K. Mis-identification of factor
Bonar RA, Kitchen DP, Kershaw G, Lippi McMillan CW, Pool JG, Shapiro SS, Schul- inhibitors by diagnostic haemostasis labora-
G. Problems and solutions in laboratory man NR, van Eys J. A more uniform tories: recognition of pitfalls and elucidation
testing for hemophilia. Semin Thromb measurement of factor VIII inhibitors. of strategies. A follow up to a large multicen-
Hemost 2013;39:816–33. Thromb Diath Haemorrh 1975;34:869–72. tre evaluation. Pathology 2007;39:504–11.
2. Key NS. Inhibitors in congenital coagulation 8. Verbrugen B, Novakova I, Wessels H, Boez- 12. Favaloro EJ, Bonar R, Duncan E, Earl G,
disorders. Br J Haematol 2005;127:379–91. man J, van den Berg M, Mauser-Bunscho- Low J, Aboud M, Just S, Sioufi J, Street A,
3. Collins PW, Chalmers E, Hart DP, Liesner ten E. The Nijmegen modification of the Marsden K, on behalf of the RCPA QAP in
RI, Rangarajan S, Talks K, Williams M, Hay Bethesda assay for factor VIII: C inhibitors: Haematology Haemostasis Committee. Iden-
Cr. Diagnosis and treatment of factor VIII improved specificity and reliability. Thromb tification of factor inhibitors by diagnostic
and IX inhibitors in congenital hemophilia Haemost 1995;73:247–51. haemostasis laboratories. A large multi-cen-
(4th edition). Br J Haematol 2013;160: 9. Favaloro EJ, Bonar R, Kershaw G, Moham- tre evaluation. Thromb Haemost 2006;
153–70. med S, Duncan E, Marsden K, RCPA Hae- 96:73–8.
4. Coppola A, Favaloro EJ, Tufano A, Di Min- matology QAP Haemostasis Committee. 13. Kitchen S, Jennings I, Preston FE, Kitchen
no MN, Cerbone AM, Franchini M. Laboratory identification of factor VIII DP, Woods TAL, Walker ID. Interlaboratory
Acquired inhibitors of coagulation factors: inhibitors in the real world: the experience variation in factor VIII: C inhibitor assay
part I-acquired hemophilia A. Semin from Australasia. Haemophilia 2010; results is sufficient to influence patient
Thromb Hemost 2012;38:433–46. 16:662–70. management: data from the UK National
5. Franchini M, Lippi G, Favaloro EJ. Acquired 10. Dardikh M, Meijer P, van den Meer F, External Assessment Service for blood
inhibitors of coagulation factors: part II. Se- Favaloro EJ, Verbruggen B. Acquired func- coagulation. Semin Thromb Hemost
min Thromb Hemost 2012;38:447–53. tional coagulation inhibitors: review on 2009;35:778–85.
6. Ma AD, Carrizosa D. Acquired factor VIII epidemiology, results of a wet-workshop 14. Bonar R, Favaloro EJ, Mardsen K. External
inhibitors: pathophysiology and treatment. on laboratory detection, and implications quality assessment of factor VIII inhibitor
Hematology Am Soc Hematol Educ Pro- for quality of inhibitor diagnosis. Semin assays. Semin Thromb Hemost 2013;
gram 2006;432–7. Thromb Hemost 2012;38:613–21. 39:320–6.
15. Bethel M, Adcock DM. Laboratory evalua- hemophilia A population of Canada. tion Blood Coagul Fibrinolysis 2004;
tion of prolonged APTT and PT. Lab Med Thromb Haemost 1998;79:872–6. 15:637–47.
2004;35:285–91. 23. Verbruggen B, van Heerde W, Novakova 29. Lippi G, Salvagno GL, Adcock DM, Gelati
16. Kershaw G, Orellana D. Mixing tests: diag- I, Lillicrap D, Giles A. A 4% solution of M, Guidi GC, Favaloro EJ. Right or wrong
nostic aides in the investigation of pro- bovine serum albumin may be used in sample received for coagulation testing?
longed prothrombin times and activated place of factor VIII: C deficient plasma in Tentative algorithms for detection of an
partial thromboplastin times. Semin the control sample in the Nijmegen modi- incorrect type of sample. Int J Lab Hematol
Thromb Hemost 2013;39:283–90. fication of the Bethesda factor VIII: C 2010;23:132–7.
17. Exner T. Conceptions and misconceptions inhibitor assay. Thromb Haemost 2002;88: 30. Adcock DM. Coagulation assays and antico-
in testing for lupus anticoagulant. J 362–4. agulant monitoring. Hematology Am Soc
Autoimmun 2000;15:179–83. 24. Meijer P, Verbruggen B. The between-labo- Hematol Educ Program 2012;2012:460–5.
18. Kershaw G, Favaloro EJ. Laboratory identi- ratory variation of factor VIII inhibitor 31. Adcock DM, Gosselin B, Kitchen S, Dwyre
fication of factor inhibitors: an update. testing: the experience of the external qual- D. The effect of dabigatran on select spe-
Pathology 2012;44:293–302. ity assessment program of the ECAT foun- cialty coagulation assays. Am J Clin Pathol
19. Adcock DM, Lippi G, Favaloro EJ. Quality dation. Semin Thromb Hemost 2009;35: 2013;139:102–9.
standards for sample processing, transporta- 786–93. 32. Bonar B, Favaloro EJ, Mohammed S, Pasa-
tion, and storage in hemostasis testing. 25. Keeling D, Mackie I, Moore GW, Greer IA, lic L, Sioufi J, Marsden K. The effect of
Semin Thromb Hemost 2012;38:576–85. Greaves M, British Committee for Stan- dabigatran on haemostasis tests: a compre-
20. Favaloro EJ, Bonar R, Kershaw G, Duncan E, dards in Haematology. Guidelines on the hensive assessment using in-vitro and
Sioufi J, Marsden K. Investigations from investigation and management of anti- ex-vivo samples. Pathology 2015;77:656–9.
external quality assurance programs reveal a phospholipid syndrome. Br J Haematol 33. Hawes EM, Deal AM, Funk-Adcock D,
high degree of variation in the laboratory 2012;157:47–53. Gosselin R, Jeanneret C, Cook AM, Taylor
identification of coagulation factor inhibitors. 26. Tiede A, Werwitzke S, Scharf RE. Labora- JM, Whinna HC, Winkler AM, Moll S. Per-
Semin Thromb Hemost 2009;35:794–805. tory diagnosis of acquired hemophilia A: formance of coagulation tests in patients on
21. Miller CH, Platt SJ, Kelly F, Soucie JM, The limitations, consequences, and challenges. therapeutic doses of dabigatran: a cross-sec-
Hemophilia Inhibitor Research Study Inves- Semin Thromb Hemost 2014;40:803–11. tional pharmacodynamics study based on
tigators. Validation of the Nijmegen–Beth- 27. Blanco AN, Cardozo MA, Sanfarelli MT, peak and trough levels. J Thromb Haemost
esda modifications to allow inhibitor Perez Bianco R, Lazzari MA. Anti-factor 2013;11:1493–502.
measurement during replacement therapy VIII inhibitors and lupus anticoagulants in 34. Francart S, Hawes EM, Deal A, Adcock D,
and facilitate inhibitor surveillance. J haemophilia A patients. Thromb Haemost Gosselin R, Jeanneret C, Friedman K, Moll
Thromb Haemost 2012;10:1055–61. 1997;77:656–9. S. Performance of coagulation tests in
22. Giles A, Verbruggen B, Rivard G, Teitel J, 28. Favaloro EJ, Posen J, Ramakrishna R, Sol- patients on therapeutic doses of rivarox-
Walker I. A detailed comparison of the per- tani S, McRae S, Just S, Aboud D, Low J, aban. A cross-sectional pharmacody-
formance of the standard versus the Nijme- Gemmel R, Kershaw G, Coleman R, Dean namic study based on peak and trough
gan modification of the Bethesda assay in M. Factor V inhibitors: rare or not so plasma levels. Thromb Haemost 2014;111:
detecting factor VIII: C inhibitors in the uncommon? A multi-laboratory investiga- 1133–40.

Int J Lab Hematology - 2015 - Adcock - Pearls and Pitfalls in Factor Inhibitor Assays

Uploaded by

Copyright:

Available Formats

Int J Lab Hematology - 2015 - Adcock - Pearls and Pitfalls in Factor Inhibitor Assays

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Int J Lab Hematology - 2015 - Adcock - Pearls and Pitfalls in Factor Inhibitor Assays

Uploaded by

Copyright:

Available Formats

International Journal of Laboratory Hematology

The Official journal of the International Society for Laboratory Hematology

REVIEW INTERNAT IONAL JOURNAL OF LABORATO RY HEMATO LOGY

Pearls and pitfalls in factor inhibitor assays

Received 2 January 2015;

tion of coagulation test(s), although not all antibodies

improvement demonstrated over the past decade [1,

ples prior to inhibitor testing. False identification of

FVIII Bethesda assay calculation sheet

Patient tube FVIII

Smith 1 0 0 0 45 0 ncalc ncalc ncalc

FIX Bethesda assay calculation sheet

Patient tube Control factor Log residual BU 9 dilution

Jones 1 0 53 0 ncalc ncalc ncalc

ncalc, unable to calculate; RA, residual activity; BU, Bethesda unit.

FIX Bethesda assay calculation sheet

Patient tube FIX

Jones 1 0 0 0 53 ncalc ncalc ncalc ncalc

as only FVIII inhibitors demonstrate time and tempera-

High-titer factor inhibitors may interfere with factor

Tests reported false identification of an inhibitor by

Laboratory error/analytical issue S A F E G UA R D S TO R E P O RT I N G FA L S E -

You might also like