AJCP  / Original Article
HITTING the Diagnosis
Testing for Heparin-Induced Thrombocytopenia in
Cancer Patients
Maly Fenelus, MD, MPH, and Ellinor I.B. Peerschke, PhD
From the Department of Laboratory Medicine, Memorial Sloan-Kettering Cancer Center, New York, NY.
Key Words: 4Ts score; Cancer; Laboratory screening; Cost savings; Coagulation; Hematology; Heparin-induced thrombocytopenia;
Management/administration; HIT GTI-PF4 ELISA; Thrombosis
Am J Clin Pathol August 2018;150:116-120
DOI: 10.1093/AJCP/AQY040
ABSTRACT
Objectives:  To evaluate the use of a pretest probability
score (4Ts score) in cancer patients to guide ordering
of laboratory screening tests for heparin-induced
thrombocytopenia (HIT).
Methods:  A retrospective chart review was conducted for
patients (n = 140) in whom laboratory testing for HIT
was requested. 4Ts scores were calculated and correlated
with heparin-endogenous platelet factor 4 antibody
enzyme-linked immunosorbent assay (ELISA) test
results.
Results:  All patients with a high pretest probability
of HIT (4Ts score = 6-7) had positive ELISA results,
compared to 26.1% of patients with intermediate (4Ts
score = 4-5) and 4.3% of patients with low (4Ts score ≤3)
pretest probability. No patients with 4Ts scores of 2 or less
had positive ELISA results.
Conclusions:  HIT can be ruled out in cancer patients
(negative predictive value and sensitivity = 100%) with
low pretest probability, defined by 4Ts scores of 2 or less,
significantly reducing the need for laboratory testing in
this patient population.
116 Am J Clin Pathol 2018;150:116-120
DOI: 10.1093/ajcp/aqy040
Downloaded from https://academic.oup.com/ajcp/article-abstract/150/2/116/5036799
by guest
on 12 July 2018
Heparin-induced thrombocytopenia (HIT) is a poten-
tially lethal complication of therapy with unfractionated hep-
arin or its derivatives.1,2 HIT is caused by the development of
antibodies directed against a molecular complex formed when
heparin associates with endogenous platelet factor 4 (PF4).
The formation of the heparin-PF4 immune complex leads to
activation of platelets, monocytes, and endothelial cells with
subsequent release of procoagulant proteins and tissue fac-
tor.1,3-5 The incidence of HIT has been reported to range from
0.1% to 5% in hospitalized patients exposed to unfractionated
heparin.2,6 HIT is a hypercoagulable state that can cause severe
venous and arterial thromboembolic complications in patients.
These complications may be associated with significant mor-
bidity and even mortality. Mortality has been reported as high
as 30% in patients with HIT and HIT thrombosis.1
The diagnosis of HIT is based on clinical assessment
and laboratory documentation of heparin-dependent
antibodies. In patients exposed to heparin, immune reac-
tions to heparin-PF4 complexes are relatively common,
varying from 8% to 50%, whereas clinical complications
are rare at 0.2% to 3%.7 To avoid overdiagnosis of HIT,
various scoring systems have been developed to help
assess the pretest probability of HIT in patients. These
scoring systems include the 4Ts score, the Lillo-Le Louёt
model, and the HIT expert probability score.8,9 The 4Ts
score is a well-validated method used as a standardized
approach to determine the pretest probability of HIT.2,10
This score is especially useful to exclude HIT given its
high negative predictive value of 99.8%.10,11
Laboratory tests to help diagnose HIT include
enzyme-linked immunosorbent assays (ELISA) that detect
© American Society for Clinical Pathology, 2018. All rights reserved. For
permissions, please e-mail: journals.permissions@oup.com

heparin-PF4 antibodies and platelet function assays such as
the serotonin release assay (SRA), which is considered the
gold standard in HIT testing. The SRA measures the ability
of heparin-PF4 immune complexes to activate normal donor
washed platelets that were exposed to 14C-serotonin.2,12
Laboratory screening tests such as the ELISA are highly
sensitive to the presence of antibodies recognizing hepa-
rin-PF4 complexes, but their lack of specificity may cause
false-positive test results that require further investigation.
An optical density (OD) of 0.4 or greater is considered a
positive ELISA result. The degree of positivity correlates
with the likelihood of HIT in patients. A  patient with an
OD of less than 1.0 is less likely to develop HIT.13,14 Using
the 4Ts scoring system to assess pretest probability can
reduce the incidence of false-positive screening test results.
In this study, we evaluated the utility of the 4Ts scor-
ing system in a cancer patient population to determine
whether a 4Ts score could be used to guide subsequent
ordering of HIT laboratory screening tests by clinicians.
There is a paucity of information focusing on HIT in
cancer patients. Prandoni et al15 found that patients with
cancer develop HIT more frequently than those without
cancer. Given that cancer patients often have thrombocy-
topenia from various causes (eg, malignancy itself, che-
motherapy, antibiotic therapy, and hematopoietic stem
cell transplantation), we aimed to identify a 4Ts score
cutoff value that would enable clinicians to determine if
laboratory testing for HIT was indicated and optimize
utilization of laboratory screening tests.
Materials and Methods
Data Collection
A retrospective chart review was conducted for all
patients with laboratory testing for HIT performed from
❚Table 1❚ 
Estimating the Pretest Probability of Heparin-Induced Thrombocytopenia: the Four Ts7,a
2 1
9
Thrombocytopenia >50% fall or nadir 20-100 × 10 /L
Timing of platelet count fall Onset between days 5 and 10 or
<1 day of recent heparin exposure
(past 100 days)
Thrombosis or other sequelae New thrombosis, skin necrosis, post
heparin bolus acute systemic reaction erythematous skin lesions, suspected
Other cause for None evident
thrombocytopenia
a
Pretest probability score: 6-8 = high, 4-5 = intermediate, 0-3 = low.
b
0, 1, or 2 for each of the four parameters.
© American Society for Clinical Pathology
Downloaded from https://academic.oup.com/ajcp/article-abstract/150/2/116/5036799
by guest
on 12 July 2018
AJCP  / Original Article
February 2015 to October 2015 and February 2016 to
March 2016. Study patients included hospitalized and
clinic patients at Memorial Sloan Kettering Cancer Center.
Individual 4Ts scores were calculated by a laboratory med-
icine fellow from clinical information obtained during
chart review of the patients’ medical records. The 4Ts score
is based on four parameters, including: (1) the degree of
thrombocytopenia, (2) the timing of platelet count decline
relative to heparin exposure, (3) the presence of thrombo-
sis, and (4) the absence of other explanations for thrombo-
cytopenia. Each parameter is given a point value from 0
to 2, and points are added for a maximum possible score
of 8 ❚Table  1❚.7 The 4Ts score is then used to assign pre-
test probability of HIT, with scores representing high (6-8),
intermediate (4-5), or low (0-3) pretest probability.7,16 Each
patient’s 4Ts score was correlated with results of the corre-
sponding heparin-PF4 IgG ELISA screening test.
Heparin-PF4 Antibody Test
Heparin-PF4 HIT antibodies (IgG) were detected
using the GTI-PF4 ELISA Kit (GTI Diagnostics,
Waukesha, WI) according to the manufacturer’s guide-
lines, using an OD cutoff value of 0.4 units. The hep-
arin-PF4 ELISA includes a confirmatory test on all
samples run using excess heparin to increase specificity.
At our institution a positive test result is reported when
the OD value is 0.4 or greater and inhibition by hepa-
rin exceeds 50%. A test result is interpreted as equivocal
when the OD value is less than 0.8 and/or inhibition by
heparin is less than 50%. Confirmatory testing by SRA
is performed at the discretion of the ordering health care
provider.
Statistics
Statistical analyses were performed using GraphPad
Prism version 6.00 for Windows (GraphPad Software, La
Pointsb
0
30%-50% fall or platelet nadir <30% or platelet nadir
10-19 × 109/L <10 × 109/L
Onset of >10 days of recent heparin Onset without recent
exposure heparin exposure
Progressive or recurrent thrombosis, None
thrombosis
Possible Definite
Am J Clin Pathol 2018;150:116-120 117
DOI: 10.1093/ajcp/aqy040
 Fenelus and Peerschke / HIT in Cancer Patients
Jolla, CA). Fisher exact test was used to detect differences
between categorical variables. A P value less than  .05 was
considered statistically significant.
Results
The 4Ts scores were correlated with HIT ELISA
screening test results in a total of 140 patients. Study
patient demographics and diagnoses are summarized in
❚Table 2❚. There was a roughly equal distribution of male
and female patients, and most patients had solid organ
malignancies. The 4Ts score distribution for all patients is
shown in ❚Figure 1❚. Most patients (82.1%) had 4Ts scores
suggesting a low probability of HIT (4Ts score  =  0-3),
whereas only 16.4% and 1.4% of patients had 4Ts scores
suggesting an intermediate (4Ts score = 4-5) or high (4Ts
score = 6-8) pretest probability of HIT, respectively.
Additionally, shown in ❚Figure  2❚ is the correlation
between 4Ts scores and heparin-PF4 IgG ELISA screen-
ing test results. A positive screening test result was defined
by an OD of 0.4 units or greater, according to manufac-
turer’s instructions. HIT ELISA test results were negative
(OD < 0.4) in 90.7% (n = 127) of patients. Each of two
patients (2/2) with a high pretest probability of HIT (4Ts
scores = 6 and 7) had a positive ELISA test result (OD
units 1.193 and 3.025, respectively). In comparison, 26.1%
(6/23) of patients with an intermediate pretest probability
❚Figure 1❚  Distribution and correlation of 4Ts scores and
heparin-induced thrombocytopenia (HIT) enzyme-linked
immunosorbent assay (ELISA) screening test results.
Pretest probability of low, intermediate, and high risk for HIT
correlate with 4Ts score of 0-3, 4-5, and 6-7, respectively.2,10
Positive and negative screening test results are indicated
based on an HIT ELISA result cutoff of <0.4 optical density
(OD) units.
118 Am J Clin Pathol 2018;150:116-120
DOI: 10.1093/ajcp/aqy040
Downloaded from https://academic.oup.com/ajcp/article-abstract/150/2/116/5036799
by guest
on 12 July 2018
❚Table 2❚ 
Summary of Patient Demographics and Malignancy Type (n = 140)
Variable Result
Age, y 61.2 ± 15.1
Gender, No. (%)
 Female 75 (53.6)
 Male 65 (46.4)
Tumor type, No. (%)
  Solid organ tumor 116 (82.9)
  Hematologic malignancy 20 (14.3)
 Othera 4 (2.9)
a
Other included patients with polycystic kidney disease, benign granulomatous
upper airway lesion, interstitial lung disease, and a patient with concurrent germ
cell tumor and chronic myelomonocytic leukemia.
of HIT (4Ts score  =  4 and 5)  had positive ELISA test
results (OD units ranging from 0.698 to 2.806). Only 4.3%
(5/115) of patients with a low pretest probability of HIT
(4Ts score ≤3) had positive ELISA test results (OD units
ranging from 0.418 to 2.506). Of note, no patients with
a 4Ts score of 2 or less had a positive HIT ELISA test
result.
Discussion
HIT is a potentially lethal complication of therapy
with unfractionated heparin or its derivatives.1,2 Although
many studies have been published on HIT, there is a pau-
city of literature focusing on cancer patients. Our findings
help fill a void in the literature by suggesting a new low
probability 4Ts score for HIT in cancer patients. A redef-
inition of the HIT 4Ts scoring is indicated in our cancer
❚Figure 2❚  Correlation between the heparin-induced throm-
bocytopenia (HIT) enzyme-linked immunosorbent assay
(ELISA) screening test results (optical density [OD] units)
and 4Ts scores depicted in a box and whiskers plot of aver-
age patient OD results by 4Ts score. Diamond symbols rep-
resent the median OD.
© American Society for Clinical Pathology
 AJCP  / Original Article

❚Table 3❚ 
Contingency Table Evaluating the Performance Characteristics of the Proposed and Traditional Low-Probability HIT Classification in
Cancer Patients
Pretest Probability of HIT
Low Intermediate/High
HIT ELISA Test Proposed 4Ts Score Traditional 4Ts Score Proposed 4Ts Score Traditional 4Ts Score
Results Cutoff ≤2 Cutoff ≤3 Cutoff ≥3 Cutoff ≥4 Total
Negative 67 110 60 17 127
Positive 0 5 13 8 13
Total 67 115 73 25 140

ELISA, enzyme-linked immunosorbent assay; HIT, heparin-induced thrombocytopenia.

patients as many are undergoing treatment regimens that
affect platelet counts, lowering the maximum possible 4Ts
score from 8 to 6.  We aimed to set up a 4Ts score cut-
off value that would allow for better utilization of HIT
ELISA testing and eliminate unnecessary testing.
These data suggest that low probability for HIT in
cancer patients may be defined by a 4Ts score of 0 to 2,
compared to the traditional cutoff 4Ts score of 0 to 3
❚Table 3❚. Based on the current patient cohort, this newly
defined low-risk pretest probability 4Ts score rules out a
positive ELISA test result with a high negative predictive
value of 100% (95% confidence interval [CI], 75%-100%)
and a sensitivity of 100% (95% CI, 95%-100%) ❚Table 4❚,
thereby eliminating subsequent laboratory testing in
these individuals. In this study population, restricting
HIT ELISA screening to cancer patients with 4Ts scores
greater than 2 would have reduced laboratory testing by
47.9% (67/140), achieving substantial conservation of
clinical and laboratory resources. At our institution, these
findings serve as an educational tool for clinicians.
Our observation that a 4Ts cutoff score of 2 or
below was consistently associated with a negative HIT
antibody titer (OD  <  0.4), and may be used safely to
discourage laboratory screening for HIT by high-sen-
sitivity/low-specificity ELISA tests. Additionally, one
year follow-up chart review showed that no patients
with screening ELISA test results of OD 0.4 or greater
and less than 0.8 were considered to have HIT clini-
cally. This finding is supported by a recently published
❚Table 4❚  and conservation of clinical laboratory resources. In
Comparison of Performance Characteristics of Proposed and
Traditional 4Ts Score Cutoff Corresponding to Low Probability
of Heparin-Induced Thrombocytopenia
Proposed 4Ts Score Traditional 4Ts Score
Cutoff ≤2, % Cutoff ≤3, %
Sensitivity 100 96
Negative predictive 100 62
value
Specificity 18 32
Positive predictive 43 87
value
study by Wong et  al,17 who recommended an optimal
cutoff 4Ts score for HIT diagnosis in cancer patients
of 5 instead of the traditionally recommended score
of greater than 3, as well as an optimal cutoff hepa-
rin-PF4 polyclonal ELISA OD of 1.004.17
Given the retrospective nature of the present
study, positive HIT screening test results could not be
confirmed with functional assays, such as serotonin
release. This limits revision of the 4Ts scoring system to
associations with negative ELISA screening test results.
Additional limitations include few positive cases and
limited duration of data collection, which correlates
with the dates the laboratory medicine fellow was on
the coagulation service.
At our institution, a significant number of ELISA
screening test orders to detect heparin-PF4 IgG are for
patients with low calculated 4Ts scores (≤2). This phenom-
enon has been described in previous studies in noncancer
patients.10,18 Utilizing a revised 4Ts scoring system has the
potential to reduce laboratory screening in approximately
50% of our patients and to provide a rapid clinical rule-out
for HIT.
Unnecessary laboratory testing for HIT is associated
with an increase in potential false-positive test results
that drives needless changes in anticoagulation therapy
and increases length of stay for patients, all contributing
to increased health care costs. Additionally, inappropri-
ate use of laboratory resources, including technologist
time and reagents, drains resources. Implementation of
pretest 4Ts scoring would enhance cancer patient care
the accountable care era, it is important to establish and
implement appropriate clinical guidelines to eliminate
unnecessary laboratory testing.
Corresponding author: Maly Fenelus, MD, MPH, Dept of
Laboratory Medicine, Memorial Sloan-Kettering Cancer Center,
327 E 64th St, New York, NY 10065; fenelusm@mskcc.org.
This work was supported in part by NIH/NCI Cancer Center
Support Grant P30 CA008748.

© American Society for Clinical Pathology Am J Clin Pathol 2018;150:116-120 119

Downloaded from https://academic.oup.com/ajcp/article-abstract/150/2/116/5036799 DOI: 10.1093/ajcp/aqy040
by guest
on 12 July 2018
 Fenelus and Peerschke / HIT in Cancer Patients
References
1. Salter BS, Weiner MM, Trinh MA, et al. Heparin-induced
thrombocytopenia: a comprehensive clinical review. J Am Coll
Cardiol. 2016;67:2519-2532.
2. Hasan M, Malalur P, Agastya M, et al. A high-value cost con-
scious approach to minimize heparin induced thrombocyto-
penia antibody (hitab) testing using the 4T score. J Thromb
Thrombolysis. 2016;42:441-446.
3. Rauova L, Zhai L, Kowalska MA, et al. Role of platelet surface
PF4 antigenic complexes in heparin-induced thrombocyto-
penia pathogenesis: diagnostic and therapeutic implications.
Blood. 2006;107:2346-2353.
4. Visentin GP, Ford SE, Scott JP, et al. Antibodies from patients
with heparin-induced thrombocytopenia/thrombosis are spe-
cific for platelet factor 4 complexed with heparin or bound to
endothelial cells. J Clin Invest. 1994;93:81-88.
5. Cines DB, Tomaski A, Tannenbaum S. Immune endotheli-
al-cell injury in heparin-associated thrombocytopenia. N Engl J
Med. 1987;316:581-589.
6. Lu BY, Kudlowitz D, Gardner LB. Clinical and labo-
ratory characteristics associated with a high optical
density anti-platelet factor 4 ELISA test. J Blood Med.
2015;6:277-283.
7. Warkentin TE, Heddle NM. Laboratory diagnosis of immune
heparin-induced thrombocytopenia. Curr Hematol Rep.
2003;2:148-157.
8. Lillo-Le Louët A, Boutouyrie P, Alhenc-Gelas M, et al.
Diagnostic score for heparin-induced thrombocytope-
nia after cardiopulmonary bypass. J Thromb Haemost.
2004;2:1882-1888.
120 Am J Clin Pathol 2018;150:116-120
DOI: 10.1093/ajcp/aqy040
Downloaded from https://academic.oup.com/ajcp/article-abstract/150/2/116/5036799
by guest
on 12 July 2018
9. Cuker A, Arepally G, Crowther MA, et al. The HIT expert
probability (HEP) score: a novel pre-test probability model for
heparin-induced thrombocytopenia based on broad expert
opinion. J Thromb Haemost. 2010;8:2642-2650.
10. Chaturvedi S, Kohli R, McCrae K. Over-testing for heparin
induced thrombocytopenia in hospitalized patients. J Thromb
Thrombolysis. 2015;40:12-16.
11. Cuker A, Gimotty PA, Crowther MA, et al. Predictive value of
the 4ts scoring system for heparin-induced thrombocytopenia: a
systematic review and meta-analysis. Blood. 2012;120:4160-4167.
12. Warkentin TE, Arnold DM, Nazi I, et al. The platelet sero-
tonin-release assay. Am J Hematol. 2015;90:564-572.
13. Chan CM, Woods CJ, Warkentin TE, et al. The role for opti-
cal density in heparin-induced thrombocytopenia: a cohort
study. Chest. 2015;148:55-61.
14. Warkentin TE, Sheppard JI, Moore JC, et al. Quantitative inter-
pretation of optical density measurements using PF4-dependent
enzyme-immunoassays. J Thromb Haemost. 2008;6:1304-1312.
15. Prandoni P, Falanga A, Piccioli A. Cancer, thrombosis
and heparin-induced thrombocytopenia. Thromb Res.
2007;120(suppl 2):S137-S140.
16. Warkentin TE. Heparin-induced thrombocytopenia: patho-
genesis and management. Br J Haematol. 2003;121:535-555.
17. Wong M, Oo TH, Qiao W, et al. Performance of 4T score
and heparin-platelet factor 4 antibody in the diagnosis of
heparin-induced thrombocytopenia (HIT) in cancer. J Thromb
Thrombolysis. 2017;44:261-266.
18. Berry CA, Nitsos I, Hillman NH, et al. Interleukin-1 in lipo-
polysaccharide induced chorioamnionitis in the fetal sheep.
Reprod Sci. 2011;18:1092-1102.
© American Society for Clinical Pathology