Abstract
Background: Cancer antigen 125 (CA125) is the best
known single tumor marker for ovarian cancer (OC). We
investigated whether the additional information of the human
epididymis protein 4 (HE4) improves diagnostic accuracy.
Methods
We retrospectively analyzed preoperative sera of
109 healthy women, 285 patients with benign ovarian masses
(cystadenoma: ns78, leimyoma: ns66, endometriosis: ns
52, functional ovarian cysts: n s79, other: n s10), 16 low
malignant potential (LMP) ovarian tumors and 125 OC
(stage I: 22, II: 15, III: 78, IV: 10). CA125 was analyzed
using the ARCHITECT system, HE4 using the ARCHI-
TECT(a) system and EIA(e) technology additionally.
Results
The lowest concentrations of CA125 and HE4 were
observed in healthy individuals, followed by patients with
benign adnexal masses and patients with LMP tumors and
OC. The area under the curve (AUC) for the differential
diagnosis of adnexal masses of CA125 alone was not sig-
nificantly different to HE4 alone in premenopausal (CA125:
86.7, HE4(a): 82.6, HE4(e): 81.6% p)0.05) but significantly
different in postmenopausal wCA125: 93.4 vs. HE4(a): 88.3
ps0.023 and vs. HE4(e): 87.8% p s0.012x patients. For
stage I OC, HE4 as a single marker was superior to CA125,
which was the best single marker in stage II-IV . The com-
bination of CA125 and HE4 using risk of malignancy algo-
rithm (ROMA) gained the highest sensitivity at 95%
specificity for the differential diagnosis of adnexal masses
wCA125: 70.9, HE4(a): 67.4, HE4(e): 66.0, ROMA(a): 76.6
*Corresponding author: Miriam Lenhard, MD, Department of
Obstetrics and Gynecology, Ludwig-Maximilians-University
Munich, Grosshadern Campus Marchioninistrasse 15, 81377
Munich, Germany
Phone: q49-89-7095-0, Fax: q49-89-7095-6724,
E-mail:
[email protected]
Received April 18, 2011; accepted August 8, 2011;
previously published online September 16, 2011
and ROMA(e): 74.5% x, especially in stage I OC wCA125:
27.3, HE4(a): 40.9, HE4(e): 40.9, ROMA(a): 45.5 and
ROMA(e): 45.5%x.
Conclusions
CA125 is still the best single marker in the
diagnosis of OC. HE4 alone and even more the combined
analysis of CA125 and HE4 using ROMA improve the diag-
nostic accuracy of adnexal masses, especially in early OC.
Keywords
CA125; human epididymis protein 4 (HE4);
ovarian cancer; risk of malignancy algorithm (ROMA);
tumor marker.
Introduction
Ovarian cancer (OC) is the leading cause of death among
gynecologic malignancies. Primary treatment includes oper-
ative cytoreduction and subsequent platinum-based com-
bined chemotherapy. Reported response rates to standard
primary treatment range around 80%, but 60%–70% of
patients with OC relapse or die within 5 years after diagnosis
(1–3). Due to the lack of diagnostic tools for early detection
of OC, the vast majority of patients are detected at a pro-
gressed stage of disease (4). Only 25%–30% of all OC
patients are diagnosed at an early stage and have better sur-
vival rates (5, 6). Persistent ovarian masses at subsequent
vaginal sonographies are difficult to handle in clinical rou-
tines. To rule out a malignant mass, operative exploration is
frequently recommended to the patient. Although most
lesions turn out to be benign, some are OCs (7). The pro-
portion of malignant ovarian tumors is higher in postmeno-
pausal in comparison to premenopausal women.
Besides vaginal sonography, tumor marker values aid in
deciding which patients to operate on. Cancer antigen 125
(CA125) as a single marker has been shown to be elevated
in the majority of patients with OC. Moreover, its level is
known to be related to stage and histological type. Still, the
diagnosis and differential diagnosis remains difficult as 20%
of all OC patients present with negative CA125 and a high
percentage of patients with benign diseases show increased
tumor marker values (8, 9). So far, the combination of
CA125 and the human epididymis protein (HE4) measured
by enzyme immunometric assay (EIA) technology (Fujirebio
Diagnostics AB, Sweden) has been described to be of addi-
tive value in the differential diagnosis of pelvic masses (10,
11). The aim of this study was to assess whether the recently
developed automated test for HE4 on the ARCHITECT sys-
tem (Abbott Diagnostics, USA) leads to comparable clinical
Results
in combination with CA125.
Bereitgestellt von | Universitaetsbibliothek der LMU Muenchen
Angemeldet | 129.187.254.47
Heruntergeladen am | 31.10.13 15:57
2082 Lenhard et al.: Diagnostic accuracy of HE4 testing systems in ovarian cancer
Article in press - uncorrected proof
Materials and methods
Patients
Women diagnosed and treated for low malignant potential (LMP)
tumors of the ovary or OC between 1985 and 2008 were included
in the study population. In the retrospective evaluation, tumor mark-
er levels were analyzed with reference to patient characteristics and
clinical data including menopausal status, histology or tumor stage.
Healthy women and patients treated for benign ovarian masses
served as controls. The study was approved by the Local Ethical
Committee. Informed consent was obtained from all patients partic-
ipating in this study.
Serum analysis
All serum samples had been obtained preoperatively at primary
diagnosis and had been stored at –80 8C. CA125 and HE4 were
analyzed in parallel using the ARCHITECT system wAbbott,
ARCHITECT, Abbott Park, IL, USA (a) x and using the EIA tech-
nology (Fujirebio Diagnostics AB, Sweden, (e) x according to the
manufacturers’ instructions.
The ARCHITECT assay for HE4 and CA125 are both two-step
immunoassays. In the first step, sample and OC 125 coated or 2H5
anti-HE4 coated paramagnetic microparticles were combined. OC
125 defined antigen sample or HE4 antigen then bind to the OC
125 coated or anti-HE4 coated microparticles. In the second step,
M11 acridinium-labeled conjugate was added after washing in the
CA125 assay and 3D8 anti-HE4 acridinium labeled conjugate in the
HE4 assay. Chemilumniscent reaction was measured by relative
light units and directly reflects CA125 and HE4 concentrations in
the serum samples.
The HE4 EIA is a solid-phase, non-competitive immunoassay
using the same 2H5 and 3D8 antibodies as described for the
ARCHITECT assay. In short, samples are incubated with biotiny-
lated anti-HE4 monoclonal antibody (MAb) 2H5 in streptavidin
coated microstrips. After washing incubation with HRP , labeled
anti-HE4 MAb 3D8 was performed. Color intensity was measured
to calculate HE4 serum concentration.
Menopausal stage was defined according to follicle-stimulating
hormone level concentrations measured. To assess the diagnostic
value of HE4 and CA125 in combination, the final ROMA was
used as described elsewhere (11):
Premenopausal patients:
Predictive Index (PI) s–12.0 q2.38=LNwHE4xq0.0626=LN
wCA125x
Postmenopausal patients:
Predictive Index (PI) s–8.09 q1.04=LNwHE4xq0.732=LN
wCA125x
Risk of malignancysexp(PI)/w1qexp(PI)x=100.
Histology
All patients were operated on by experienced gynecological sur-
geons. Tumor typing and staging were performed by experienced
gynecological pathologists according to the criteria of the Interna-
tional Federation of Gynaecologists and Obstetricians (FIGO) and
the International Union against Cancer (IUCC). Histology, FIGO
stage and tumor grade according to the World Health Organization
(WHO) were recorded.
Statistics
Statistical analysis was performed using SAS V9.2 (SAS Institute
Inc., Cary NC, USA). Tumor marker results are given as median,
range and percentile. Differences between groups were evaluated
using the non-parametric Wilcoxon test (e.g., histological subtype)
or the non-parametric Jonckheere-Terpstra test (e.g., tumor stage).
The Passing and Bablok regression was performed to test for the
equality of measurements of the two different analytical HE4 tech-
nologies (12). Moreover, the Spearman correlation coefficient was
calculated for the values obtained with the HE4 EIA and HE4
ARCHITECT systems.
The AUCs were determined to reflect the relationship between
sensitivity and specificity for single tumor markers or tumor marker
combinations. To compare AUCs for statistical significance, p-val-
ues were calculated (13). Sensitivities were also calculated at set
specificities of 75%, 90% and 95% for each marker and marker com-
binations. p-Values of-0.05 were regarded as statistically significant.
Results
Patient characteristics
We retrospectively analyzed preoperative sera of 109 healthy
women, 285 patients with benign ovarian masses, 16 LMP
tumors of the ovary and 125 OC. Patient’s characteristics
including median age, menopausal status, histology, stage
and grade are shown in detail in Table 1.
Tumor marker values
The Passing and Bablok regression analysis indicated com-
parable results for the two HE4 technologies (all values:
P/B regression HE4(a) s1.024=HE4(e)q6.030; n s646).
There was a good correlation between HE4 EIA and HE4
ARCHITECT measurement (Spearman: rs0.91) (Figure 1).
The lowest median concentrations of CA125 and HE4
were observed in healthy individuals, followed by patients
with benign disorders and patients presenting with LMP
tumors. The highest marker concentrations were observed in
OC patients. For each sub-group, the two HE4 technologies
showed similar values within the group (Table 2).
For the histological type, CA125 showed the highest
serum concentrations in serous OC patients. In general, HE4
measured by ARCHITECT and EIA showed comparable
Results
within each histological group of OC patients. The
highest median HE4 release was found for serous OCs, fol-
lowed by endometrioid tumors. Significantly different mark-
er levels between histological subtypes were observed
between serous and mucinous OC in CA125 (ps0.0011) and
HE4 ARCHITECT (p s0.0029) as well as HE4 EIA
(ps0.0026) and also comparing serous to endometrioid OC
in CA125 levels (p s0.0063) (Table 2).
With regard to tumor stage, the lowest median tumor
marker values were found in patients with stage I OC. Sig-
nificant different tumor marker levels with regard to tumor
stage were observed for CA125 (p -0.001), HE4 ARCHI-
TECT (p-0.0098) and HE4 EIA (p -0.0089). Again, the
serum testing for HE4 with the ARCHITECT or EIA tech-
nology showed comparable results within each group of
patients (Table 3).
ROC curves were calculated for the comparison of
patients with benign ovarian masses vs. OC and LMP tumor
Bereitgestellt von | Universitaetsbibliothek der LMU Muenchen
Angemeldet | 129.187.254.47
Heruntergeladen am | 31.10.13 15:57
Lenhard et al.: Diagnostic accuracy of HE4 testing systems in ovarian cancer 2083
Article in press - uncorrected proof
Table 1 Patients’ characteristics for healthy women, patients with benign ovarian masses, LMP or OC including median age, menopausal
status, histology, stage and grade.
Patients Number (n) Median age Menopausal Histology (n) Stage Grade
(range) status (n) (FIGO)
Healthy 109 38.4 (21.5–80.0) Premenopausal: 81
women Postmenopausal: 20
NA: 8
Benign 285 44.6 (18.5–87.3) Premenopausal: 160 Cystadenoma: 78
ovarian Postmenopausal: 125 Leiomyoma: 66
mass Endometriosis: 52
Functional ovarian cyst: 79
Inflamm. adnexal disease: 3
Teratoma: 7
LMP 16 59.7 (23.6–88.3) Premenopausal: 3 Endometrioid: 1 I: 15 G2: 2
Postmenopausal: 13 Mucinous: 7 II: 0 NA: 14
Serous: 8 III: 1
OC 125 62.9 (22.7–88.2) Premenopausal: 27 Endometroid: 12 I: 22 G1: 8
Postmenopausal: 98 Mucinous: 8 II: 15 G2: 30
Serous: 84 III: 78 G3: 68
Other: 21 IV: 10 NA: 19
Figure 1 The Passing and Bablok regression for the two HE4
technologies (all values: P/B regression Y s1.024=Xq6.030;
ns646; Spearman: rs0.91).
patients for the single marker HE4(a), HE4(e) and CA125
as well as ROMA using either HE4(a) or HE4(e) (Table 4).
The combined analysis of CA125 and HE4 (ROMA), either
with the ARCHITECT or EIA, was significantly superior to
both HE4 tests alone wHE4(e) vs. ROMA(e) p s0.008,
HE4(a) vs. ROMA(a) p s0.013x. No significant differences
were observed between the ROC curves for CA125 in com-
parison to HE4 or ROMA using both HE4 technologies
wCA125 vs. HE4(e) or vs. HE4(a) p )0.05, CA125 vs.
ROMA(e) or vs. ROMA(a) p )0.05x. Moreover, there were
no significant differences between both HE4 testing systems
and between both ROMA results wHE4(e) vs. HE4(a) p )
0.05, ROMA(e) vs. ROMA(a) p )0.05x (Figure 2).
At a set specificity of 95%, sensitivities were lowest for
all single markers or the marker combinations at premeno-
pausal status. When comparing single markers and ROMA
in premenopausal patients, the AUCs of CA125, HE4(e) and
HE4(a) and both ROMA were not significantly different
(p)0.05 each).
However, AUCs for both ROMA and both HE4 were sig-
nificantly different in comparison to CA125 alone in post-
menopausal patients wCA125 vs. HE4(a) p s0.023, CA125
vs. HE4(e) p s0.012, CA125 vs. ROMA(a) p s0.001,
CA125 vs. ROMA(e) p s0.001 (Table 4)x.
Table 5 shows the results for the comparison of patients
with a benign ovarian mass vs. patients with OC for different
tumor stages. Sensitivities were calculated at set specificities
of 75%, 90% and 95%, respectively. When raising specific-
ity, the highest loss in sensitivity is generally found in early
stage disease. However, the combination of both ROMA
gained the highest sensitivity at 95% specificity for stage I
OC patients, followed by HE4 alone and CA125 alone. Only
at stage I OC, CA125 seems to be inferior to both HE4 and
both ROMA (Table 5).
Discussion
OC is often detected at progressed stage and ranks as the
fifth most common cause of death in women (14). So far,
data for OC screening with vaginal sonography, clinical
examination and tumor marker evaluation are not satisfactory
(15, 16). Moreover, the interpretation of tumor marker ele-
vation remains difficult for the known variation in tumor
marker release due to inflammation, endometriosis, liver dis-
ease, ovulation or menopausal status (17, 18), which we
could also observe in our patient population.
Tumor marker CA125
There have been numerous studies describing preoperative
CA125 in OC patients and few studies for ovarian borderline
Bereitgestellt von | Universitaetsbibliothek der LMU Muenchen
Angemeldet | 129.187.254.47
Heruntergeladen am | 31.10.13 15:57
2084 Lenhard et al.: Diagnostic accuracy of HE4 testing systems in ovarian cancer
Article in press - uncorrected proof
Table 2 CA125 and HE4 (ARCHITECT and EIA) concentrations for healthy women, patients with benign ovarian masses, patients
diagnosed with LMP tumor or OC showing median, range, 5th and 95th percentile (Pctl). In addition, CA125 and HE4 (ARCHITECT and
EIA) concentrations are shown in detail for the histological subtypes. Significantly different results were observed between serous and
mucinous OC in CA125 (p s0.0011), HE4 ARCHITECT (p s0.0029) and HE4 EIA (p s0.0026) and comparing serous to endometrioid
OC in CA125 (p s0.0063).
Group n Marker Unit Median Range 5th Pctl 95th Pctl
Healthy women 109 CA 125 ARCHITECT U/mL 13.5 4.0–49.7 7.4 30.0
HE4 ARCHITECT pmol 40.4 4.5–111 15.2 72.0
HE4 EIA pmol 32.5 2.5–444 2.5 77.2
Benign ovarian mass 285 CA 125 ARCHITECT U/mL 19.6 4.0–276 8.4 124
HE4 ARCHITECT pmol 40.3 18.8–1178 25.0 94.6
HE4 EIA pmol 35.3 2.5–1120 2.5 96.7
LMP 16 CA 125 ARCHITECT U/mL 34.7 18.1–385 18.1 335
HE4 ARCHITECT pmol 52.1 28.2–399 28.2 399
HE4 EIA pmol 43.7 13.6–441 13.6 441
OC 125 CA 125 ARCHITECT U/mL 391 12.5–35813 28.3 7075
All HE4 ARCHITECT pmol 242 29.0–7507 37.7 2954
HE4 EIA pmol 246 9.6–5669 33.8 3357
Serous 84 CA 125 ARCHITECT U/mL 656.5 16.2–35813 39.4 3321
HE4 ARCHITECT pmol 386.0 31.6–7507 54.2 3124
HE4 EIA pmol 364.0 17.2–5669 40.8 3398
Mucinous 8 CA 125 ARCHITECT U/mL 126.5 29.2–285 29.2 285
HE4 ARCHITECT pmol 74.1 35.4–147 35.4 447
HE4 EIA pmol 75.6 33.8–335 33.8 335
Endometrioid 12 CA 125 ARCHITECT U/mL 70.3 21.3–1749 21.3 1749
HE4 ARCHITECT pmol 165.5 39.7–3469 39.7 3469
HE4 EIA pmol 146.0 33.7–3357 33.7 3357
Other 21 CA 125 ARCHITECT U/mL 211.0 12.5–13804 20.3 1564
HE4 ARCHITECT pmol 161.0 29.0–2301 33.0 1134
HE4 EIA pmol 146.0 9.6–1901 31.7 1202
Table 3 CA125 and HE4 (ARCHITECT and EIA) concentrations in OC patients, detailed for tumor stage I-IV , showing median, range,
5th and 95th percentile (Pctl). Significant differences between stages were observed for CA125 (two sided p -0.001), HE4 ARCHITECT
(two sided p-0.0098) and HE4 EIA (two sided p -0.0089).
Group n Marker Unit Median Range 5th Pctl 95th Pctl
All OC 125 CA125 U/mL 391 12.5–35813 28.3 7075
ARCHITECT pmol 242 29.0–7507 37.7 2954
HE4 ARCHITECT pmol 246 9.6–5669 33.8 3357
HE4 EIA
FIGO I 22 CA125 U/mL 38.2 12.5–1564 16.2 1459
ARCHITECT pmol 66.0 29.0–2301 33.0 1134
HE4 ARCHITECT pmol 71.3 9.6–1901 31.7 1100
HE4 EIA
FIGO II 15 CA125 U/mL 1519 34.6–5855 34.6 5855
ARCHITECT pmol 538 53.2–7507 53.2 7507
HE4 ARCHITECT pmol 624 43.8–5633 43.8 5633
HE4 EIA
FIGO III 78 CA125 U/mL 505 28.5–35813 56.2 9227
ARCHITECT pmol 340 33.2–5184 50.5 3124
HE4 ARCHITECT pmol 317 32.1–5669 35.8 3398
HE4 EIA
FIGO IV 10 CA125 U/mL 559 28.3–8323 28.3 8323
ARCHITECT pmol 300 31.6–1114 31.6 1114
HE4 ARCHITECT pmol 341 17.2–1076 17.2 1076
HE4 EIA
patients (19–21). OC patients, but also ovarian borderline
tumor patients are known to present with elevated CA125 in
most cases (21, 22). Our data confirm that serum CA125 in
OC patients differ in median from healthy controls, patients
with benign gynecologic disease and patients with ovarian
borderline tumors. Patients with OC generally show the high-
Bereitgestellt von | Universitaetsbibliothek der LMU Muenchen
Angemeldet | 129.187.254.47
Heruntergeladen am | 31.10.13 15:57
Lenhard et al.: Diagnostic accuracy of HE4 testing systems in ovarian cancer 2085
Article in press - uncorrected proof
Figure 2 ROC curves for CA125, HE4 wEIA and ARCHI-
TECT(a)x and ROMA using either the HE4 EIA or ARCHITECT(a)
technology.
Diagnostic accuracy is calculated for patients with benign ovarian
mass vs. all OC patients and patients with LMP tumors.
est levels of CA125, healthy women the lowest levels. More-
over, patients with progressed OC present with higher tumor
marker values, which was also described in literature (23).
Median tumor marker values differ in patients with various
histological subtypes of OC. Our data show serous OC
patients to have the highest CA125 concentrations in median,
which agrees with the data of other centers (23, 24).
Tumor marker HE4
CA125 is the most commonly used tumor marker in the diag-
nosis and differential diagnosis of pelvic masses (10, 11).
New tumor markers like HE4 are gaining importance. Drap-
kin et al. showed that HE4 is distributed in a region of the
cytoplasm with a perinuclear pattern reminiscent of the
endoplasmic reticulum and the Golgi apparatus (24). The
HE4 gene encodes a protease inhibitor with a role in protec-
tive immunity and is primarily expressed in the reproductive
tract and upper airways (24, 25). Moreover, the HE4 gene
product is N-glycosylated and secreted into the extracellular
environment of cancer cells, and elevated serum levels can
be measured in these patients. Some studies analyzed HE4
in gynecological cancer patients (26). We concentrated on
OC and ovarian borderline tumor patients with appropriate
control groups of gynecological patients and healthy women
and compared the results of two different HE4 measurement
systems. Tumor marker values for HE4 were lowest in stage
I disease with both technologies used. The data agrees with
the results published by Montagnana et al. (10), who also
found a stage dependent HE4 release in OC. Interestingly,
HE4 was more frequently expressed in early stage disease
when compared to CA125. Especially at very high specific-
ity, HE4 alone achieves 40.9% sensitivity with both HE4
techniques. In comparison to single HE4 values, CA125
alone only has a sensitivity of 27.3% and is therefore inferior
to HE4 testing. Especially at early stage disease, HE4 seems
to improve the identification of OC patients. Our results are
in accordance with the data published by Havrilesky et al. (27).
Similar to CA125, HE4 shows the highest tumor marker
values for serous OC patients, which is also in agreement
with data in literature (10). Interestingly, HE4 levels are also
relatively high in patients with endometrioid tumors. We
recently found HE4 to be less frequently elevated in benign
gynecological disease (data not presented in detail for dif-
ferent diseases here) like endometriosis, where CA125 can
attain relatively high serum values. Still, there are diseases
in which HE4 also lacks specificity, like inflammation or
renal failure (28). Moreover, recent data describe a signifi-
cant variation of HE4 release in healthy women under the
age of 35, which depends on the female hormonal cycle.
Therefore, Anastasi et al. point out that it is necessary to
properly interpret HE4 data with regard to the female hor-
monal cycle (29).
Marker combination
Various studies investigated the use of tumor marker com-
binations in OC patients (11, 27, 30–37). The combination
of CA125 and HE4 seems to be favorable in most studies
for the differential diagnosis of these patients (31, 34).
Moore et al. used the dual marker combination of HE4 and
CA125 in ROMA to classify women into high and low risk
groups (11).
In this study, we assessed whether the automated test for
HE4 with the ARCHITECT system and the manual HE4 EIA
lead to comparable clinical results in combination with CA125.
The results achieved with the dual marker combination
and information on the patient’s menopausal status, as per-
formed in ROMA, are comparable to the published result of
the most commonly used risk of malignancy index (RMI)
(38–40), RMI utilizes a combination of serum CA125 val-
ues, pelvic sonography and the menopausal status (41). In
Jacobs et al. a sensitivity of 95.1% is noted at 76.5% spec-
ificity (42). In a recently published study, Moore et al. com-
pared the RMI with ROMA (43). At a set specificity of 75%,
ROMA had a sensitivity of 94.3% and RMI had a sensitivity
of 84.6% for distinguishing benign findings from EOC (43).
Further analysis for patients with early stage disease showed
a sensitivity of 85.3% for ROMA compared to 64.7% for
RMI. The authors therefore concluded that the dual marker
algorithm utilizing HE4 and CA125 achieves a higher sen-
sitivity than RMI in the diagnosis of OC (43).
A recently published study by Montagnana et al. con-
cludes ROMA to be excellent in the detection of OC of post-
menopausal but not premenopausal women (30). This
finding goes along with our results as we observed AUCs of
83.1% with ROMA using the ARCHITECT and 82.0% using
the EIA technology in premenopausal women in comparison
to 93.9% (ARCHITECT) and 93.2% (EIA) in postmenopau-
Bereitgestellt von | Universitaetsbibliothek der LMU Muenchen
Angemeldet | 129.187.254.47
Heruntergeladen am | 31.10.13 15:57
2086 Lenhard et al.: Diagnostic accuracy of HE4 testing systems in ovarian cancer
Article in press - uncorrected proof
Table 4 Sensitivity at 95% specificity and AUC values with lower confidence limits (LCL) and upper confidence limits (UCL) in the
differentiation of benign and malignant gynecological masses for CA125, HE4 ARCHITECT, HE4 EIA and ROMA using the HE4 ARCHI-
TECT and HE4 EIA system. Results are also shown in detail by menopausal status.
Patients Marker Sensitivity at AUC %
95% specificity (LCL–UCL)
All patients: CA 125 ARCHITECT 70.9% 91.1 (88.2–94.0)
benign ovarian mass HE4 ARCHITECT 67.4% 89.0 (85.4–92.7)
(ns285) vs. OCqLMP (ns141) HE4 EIA 66.0% 88.2 (84.5–91.9)
ROMA (HE4 ARCHITECT) 76.6% 92.4 (89.3–95.6)
ROMA (HE4 EIA) 74.5% 91.8 (88.6–95.0)
Premenopausal patients: CA 125 ARCHITECT 60.0% 86.7 (78.8–94.5)
benign ovarian mass HE4 ARCHITECT 66.7% 82.6 (72.1–93.5)
(ns160) vs. OCqLMP (ns30) HE4 EIA 56.7% 81.6 (71.1–92.0)
ROMA (HE4 ARCHITECT) 66.6% 83.1 (72.6–93.7)
ROMA (HE4 EIA) 56.7% 82.0 (71.7–92.3)
postmenopausal patients: CA 125 ARCHITECT 73.0% 93.4 (90.6–96.3)
benign ovarian mass HE4 ARCHITECT 66.8% 88.3 (84.0–92.7)
(ns125) vs. OCqLMP (ns111) HE4 EIA 64.9% 87.8 (83.4–92.2)
ROMA (HE4 ARCHITECT) 72.1% 93.9 (91.1–96.7)
ROMA HE4 EIA) 71.2% 93.2 (90.2–96.3)
Table 5 Tumor marker sensitivity at set specificity of 75%, 90% and 95% for CA125, HE4 ARCHITECT, HE4 EIA, ROMA using the
HE4 ARCHITECT and ROMA using the HE4 EIA, detailed for tumor stage I–IV .
Benign Marker Sensitivity at Sensitivity at Sensitivity at
ovarian mass 75% specificity 90% specificity 95% specificity
(ns285) vs.
OC CA125 ARCHITECT 50.0% 27.3% 27.3%
FIGO I HE4 ARCHITECT 68.2% 45.5% 40.9%
(ns22) HE4 EIA 63.6% 50.0% 40.9%
ROMA (HE4 ARCHITECT) 86.4% 59.1% 45.5%
ROMA (HE4 EIA) 81.8% 59.1% 45.5%
OC CA 125 ARCHITECT 93.3% 86.7% 86.7%
FIGO II HE4 ARCHITECT 100% 80.0% 80.0%
(ns15) HE4 EIA 100% 80.0% 80.0%
ROMA (HE4 ARCHITECT) 93.3% 86.7% 86.7%
ROMA (HE4 EIA) 93.3% 86.7% 80.0%
OC CA125 ARCHITECT 98.7% 92.3% 89.7%
FIGO III HE4 ARCHITECT 94.9% 88.5% 83.3%
(ns78) HE4 EIA 92.3% 85.9% 80.8%
ROMA (HE4 ARCHITECT) 98.7% 96.2% 92.3%
ROMA (HE4 EIA) 98.7% 93.6% 89.7%
OC CA125 ARCHITECT 90.0% 80.0% 80.0%
FIGO IV HE4 ARCHITECT 80.0% 80.0% 70.0%
(ns10) HE4 EIA 80.0% 70.0% 70.0%
ROMA (HE4 ARCHITECT) 80.0% 80.0% 80.0%
ROMA (HE4 EIA) 80.0% 80.0% 80.0%
sal women. This variability should always be taken into
account when ROMA score results are compared (30).
Still, results on CA125, HE4 and ROMA are partly con-
troversial. Ruggeri et al. found HE4 to be more specific and
accurate than CA125 (34), whereas Van Gorg et al. conclud-
ed HE4 and ROMA not to increase the detection of ovarian
cancer (37). Jacob et al. analyzed HE4 and CA125 with or
without ROMA and RMI and concluded not to see a benefit
from combining HE4 and CA125 as tumor markers in a clin-
ical setting (36).
We observed the highest benefit for HE4 and ROMA in
OC patients and early stage disease. This finding was also
described by others (11, 36).
Altogether, the differences in CA125, HE4 and ROMA
Results
may be attributable to various numbers of patients,
tumor stages, menopausal status or histologic subtypes in the
ovarian cancer patient study groups and also control groups.
Strengths of this study are the large patient and control
groups, the persistent high standard of surgery by gyneco-
logic oncologists at a specialized academic institution and
Bereitgestellt von | Universitaetsbibliothek der LMU Muenchen
Angemeldet | 129.187.254.47
Heruntergeladen am | 31.10.13 15:57
Lenhard et al.: Diagnostic accuracy of HE4 testing systems in ovarian cancer 2087
Article in press - uncorrected proof
the consistent histopathologic review by expert gynecologic
oncology pathologists.
Conclusions
In the differential diagnosis of OC patients, CA125 still rep-
resents the best single tumor marker. The combined analysis
of CA125 and HE4, using either the ARCHITECT or EIA
system, improves the diagnostic accuracy in the distinction
of ovarian tumors, especially in early OC. Both HE4 testing
Methods
can be used in clinical routines since the HE4
ARCHITECT shows comparable results with the HE4 EIA
technology.
Acknowledgments
We thank Dr. Barry Dowell and Sabrina Chan (Abbott Diagnostics,
USA) und Dr. Olle Nilsson (Fujirebio Diagnostics, Sweden) for
their support with reagents and equipment and for contributing their
scientific know how in data evaluation and discussion.
Conflict of interest statement
Authors’ conflict of interest disclosure: The authors stated that
there are no conflicts of interest regarding the publication of this
article.
Research funding: None declared.
Employment or leadership: None declared.
Honorarium: None declared.
References
1. Berek JS, Bertelsen K, du Bois A, Brady MF , Carmichael J,
Eisenhauer EA, et al. Advanced epithelial ovarian cancer: 1998
consensus statements. Ann Oncol 1999;10:87–92.
2. Omura GA, Brady MF , Homesley HD, Y ordan E, Major FJ,
Buchsbaum HJ, et al. Long-term follow-up and prognostic factor
analysis in advanced ovarian carcinoma: the Gynecologic Oncol-
ogy Group experience. J Clin Oncol 1991;9:1138–50.
3. Thigpen JT, Bertelsen K, Eisenhauer EA, Hacker NF , Lund B,
Sessa C. Long-term follow-up of patients with advanced ovarian
carcinoma treated with chemotherapy. Ann Oncol 1993;4:35–
40.
4. Cannistra SA. Cancer of the ovary. N Engl J Med 1993;329:
1550–9.
5. Chan J, Fuh K, Shin J, Cheung M, Powell C, Chen LM, et al.
The treatment and outcomes of early-stage epithelial ovarian can-
cer: have we made any progress? Br J Cancer 2008;98:1191–6.
6. Lenhard SM, Bufe A, Kumper C, Stieber P , Mayr D, Hertlein L,
et al. Relapse and survival in early-stage ovarian cancer. Arch
Gynecol Obstet 2009;280:71–7.
7. Fishman DA, Cohen L, Blank SV , Shulman L, Singh D, Bozorgi
K, et al. The role of ultrasound evaluation in the detection of
early-stage epithelial ovarian cancer. Am J Obstet Gynecol
2005;192:1214–21.
8. Bast RC Jr, Klug TL, St John E, Jenison E, Niloff JM, Lazarus
H, et al. A radioimmunoassay using a monoclonal antibody to
monitor the course of epithelial ovarian cancer. N Engl J Med
1983;309:883–7.
9. Jacobs I, Bast RC Jr. The CA 125 tumour-associated antigen:
a review of the literature. Hum Reprod 1989;4:1–12.
10. Montagnana M, Lippi G, Ruzzenente O, Bresciani V , Danese
E, Scevarolli S, et al. The utility of serum human epididymis
protein 4 (HE4) in patients with a pelvic mass. J Clin Lab Anal
2009;2:331–5.
11. Moore RG, McMeekin DS, Brown AK, DiSilvestro P , Miller
MC, Allard WJ, et al. A novel multiple marker bioassay util-
izing HE4 and CA125 for the prediction of ovarian cancer in
patients with a pelvic mass. Gynecol Oncol 2009;112:40–6.
12. Passing H, Bablok W . A new biometrical procedure for testing
the equality of measurements from two different analytical
methods. Application of linear regression procedures for meth-
od comparison studies in clinical chemistry, Part I. J Clin Chem
Clin Biochem 1983;21:709–20.
13. Hanley JA, McNeil BJ. A method of comparing the areas under
receiver operating characteristic curves derived from the same
cases. Radiology 1983;148:839–43.
14. Heintz AP , Odicino F , Maisonneuve P , Quinn MA, Benedet JL,
Creasman WT, et al. Carcinoma of the ovary. FIGO 6th Annual
Report on the Results of Treatment in Gynecological Cancer.
Int J Gynaecol Obstet 2006;95:S161–92.
15. Menon U, Gentry-Maharaj A, Hallett R, Ryan A, Burnell M,
Sharma A, et al. Sensitivity and specificity of multimodal and
ultrasound screening for ovarian cancer, and stage distribution
of detected cancers: results of the prevalence screen of the UK
Collaborative Trial of Ovarian Cancer Screening (UKCTOCS).
Lancet Oncol 2009;10:327–40.
16. Joyner AB, Runowicz CD. Ovarian cancer screening and early
detection. Womens Health (Lond Engl) 2009;5:693–9.
17. Medeiros LR, Rosa DD, da Rosa MI, Bozzetti MC. Accuracy
of CA 125 in the diagnosis of ovarian tumors: a quantitative
systematic review. Eur J Obstet Gynecol Reprod Biol 2009;
142:99–105.
18. Kenemans P , Yedema CA, Bon GG, von Mensdorff-Pouilly S.
CA 125 in gynecological pathology – a review. Eur J Obstet
Gynecol Reprod Biol 1993;49:115–24.
19. Gotlieb WH, Soriano D, Achiron R, Zalel Y, Davidson B,
Kopolovic J, et al. CA 125 measurement and ultrasonography
in borderline tumors of the ovary. Am J Obstet Gynecol
2000;183:541–6.
20. van Nagell JR Jr, DePriest PD, Ueland FR, DeSimone CP , Coo-
per AL, McDonald JM, et al. Ovarian cancer screening with
annual transvaginal sonography: findings of 25,000 women
screened. Cancer 2007;109:1887–96.
21. Lenhard MS, Nehring S, Nagel D, Mayr D, Kirschenhofer A,
Hertlein L, et al. Predictive value of CA 125 and CA 72-4 in
ovarian borderline tumors. Clin Chem Lab Med 2009;47:
537–42.
22. Zurawski VR Jr, Orjaseter H, Andersen A, Jellum E. Elevated
serum CA 125 levels prior to diagnosis of ovarian neoplasia:
relevance for early detection of ovarian cancer. Int J Cancer
1988;42:677–80.
23. Y oung RC, Walton LA, Ellenberg SS, Homesley HD, Wilbanks
GD, Decker DG, et al. Adjuvant therapy in stage I and stage
II epithelial ovarian cancer. Results of two prospective random-
ized trials. N Engl J Med 1990;322:1021–7.
24. Drapkin R, von Horsten HH, Lin Y, Mok SC, Crum CP , Welch
WR, et al. Human epididymis protein 4 (HE4) is a secreted
glycoprotein that is overexpressed by serous and endometrioid
ovarian carcinomas. Cancer Res 2005;65:2162–9.
25. Bouchard D, Morisset D, Bourbonnais Y, Tremblay GM. Pro-
Bereitgestellt von | Universitaetsbibliothek der LMU Muenchen
Angemeldet | 129.187.254.47
Heruntergeladen am | 31.10.13 15:57
2088 Lenhard et al.: Diagnostic accuracy of HE4 testing systems in ovarian cancer
Article in press - uncorrected proof
teins with whey-acidic-protein motifs and cancer. Lancet Oncol
2006;7:167–74.
26. Moore RG, Brown AK, Miller MC, Badgwell D, Lu Z, Allard
WJ, et al. Utility of a novel serum tumor biomarker HE4 in
patients with endometrioid adenocarcinoma of the uterus.
Gynecol Oncol 2008;110:196–201.
27. Havrilesky LJ, Whitehead CM, Rubatt JM, Cheek RL, Groelke
J, He Q, et al. Evaluation of biomarker panels for early stage
ovarian cancer detection and monitoring for disease recurrence.
Gynecol Oncol 2008;110:374–82.
28. Dong L, Chang XH, Ye X, Zhu LR, Zhao Y, Tian L, et al. The
values of serum human epididymis secretory protein 4 and
CA(125) assay in the diagnosis of ovarian malignancy. Zhong-
hua Fu Chan Ke Za Zhi 2008;43:931–6.
29. Anastasi E, Granato T, Marchei GG, Viggiani V , Colaprisca B,
Comploj S, et al. Ovarian tumor marker HE4 is differently
expressed during the phases of the menstrual cycle in healthy
young women. Tumour Biol 2010;31:411–5.
30. Montagnana M, Danese E, Ruzzenente O, Bresciani V , Nuzzo
T, Gelati M, et al. The ROMA (Risk of Ovarian Malignancy
Algorithm) for estimating the risk of epithelial ovarian cancer
in women presenting with pelvic mass: is it really useful? Clin
Chem Lab Med 2011;49:521–5.
31. Kim YM, Whang DH, Park J, Kim SH, Lee SW , Park HA, et
al. Evaluation of the accuracy of serum human epididymis pro-
tein 4 in combination with CA125 for detecting ovarian cancer:
a prospective case-control study in a Korean population. Clin
Chem Lab Med 2011;49:527–34.
32. Bast RC Jr., Badgwell D, Lu Z, Marquez R, Rosen D, Liu J,
et al. New tumor markers: CA125 and beyond. Int J Gynecol
Cancer 2005;15:274–81.
33. Moore RG, Brown AK, Miller MC, Skates S, Allard WJ, Verch
T, et al. The use of multiple novel tumor biomarkers for the
detection of ovarian carcinoma in patients with a pelvic mass.
Gynecol Oncol 2008;108:402–8.
34. Ruggeri G, Bandiera E, Zanotti L, Belloli S, Ravaggi A, Roma-
ni C, et al. HE4 and epithelial ovarian cancer: Comparison and
clinical evaluation of two immunoassays and a combination
algorithm. Clin Chim Acta 2011;412:1447–53.
35. Park Y, Kim Y, Lee EY, Lee JH, Kim HS. Reference ranges
for HE4 and CA125 in a large Asian population by automated
assays and diagnostic performances for ovarian cancer. Int J
Cancer 2011. April 11. doi: 10.1002/ijc.26129 wEpub ahead of
printx.
36. Jacob F , Meier M, Caduff R, Goldstein D, Pochechueva T,
Hacker N, et al. No benefit from combining HE4 and CA125
as ovarian tumor markers in a clinical setting. Gynecol Oncol
2011;121:487–91.
37. Van Gorp T, Cadron I, Despierre E, Daemen A, Leunen K,
Amant F , et al. HE4 and CA125 as a diagnostic test in ovarian
cancer: prospective validation of the Risk of Ovarian Malig-
nancy Algorithm. Br J Cancer 2011;104:863–70.
38. McDonald JM, Doran S, DeSimone CP , Ueland FR, DePriest
PD, Ware RA, et al. Predicting risk of malignancy in adnexal
masses. Obstet Gynecol 2010;115:687–94.
39. Kader Ali Mohan GR, Jaaback K, Proietto A, Robertson R,
Angstetra D. Risk Malignancy Index (RMI) in patients with
abnormal pelvic mass: Comparing RMI 1, 2 and 3 in an Aus-
tralian population. Aust N Z J Obstet Gynaecol 2010;50:77–80.
40. van den Akker PA, Aalders AL, Snijders MP , Kluivers KB,
Samlal RA, Vollebergh JH, et al. Evaluation of the Risk of
Malignancy Index in daily clinical management of adnexal
masses. Gynecol Oncol 2010;116:384–8.
41. Moolthiya W , Yuenyao P . The risk of malignancy index (RMI)
in diagnosis of ovarian malignancy. Asian Pac J Cancer Prev
2009;10:865–8.
42. Jacobs I, Oram D, Fairbanks J, Turner J, Frost C, Grudzinskas
JG. A risk of malignancy index incorporating CA 125, ultra-
sound and menopausal status for the accurate preoperative diag-
nosis of ovarian cancer. Br J Obstet Gynaecol 1990;97:922–9.
43. Moore RG, Jabre-Raughley M, Brown AK, Robison KM,
Miller MC, Allard WJ, et al. Comparison of a novel multiple
marker assay vs. the Risk of Malignancy Index for the predic-
tion of epithelial ovarian cancer in patients with a pelvic mass.
Am J Obstet Gynecol 2010;203:228e1–6.
Bereitgestellt von | Universitaetsbibliothek der LMU Muenchen
Angemeldet | 129.187.254.47
Heruntergeladen am | 31.10.13 15:57
Text is read by the "Ask this paper" AI Q&A widget below.
Extraction quality varies by source — PMC NXML preserves structure
cleanly, OA-HTML may include some navigation residue, and OA-PDF can
have broken hyphenation. The publisher copy
(via DOI)
is the canonical version.