Keywords
benign adnexal mass - International Ovarian Tumour Analysis - IOTA simple rules -
malignant adnexal mass - three-dimensional ultrasound - power Doppler
Introduction
Accurate preoperative differentiation of benign and malignant adnexal masses is essential
to guide timely referral and personalized surgical planning. Clinicians use demographic
parameters, tumor markers, and ultrasound predictors to classify adnexal masses in
clinical practice.[1] Evidence-based, validated International Ovarian Tumour Analysis Simple Rules (IOTA-SR)
are the most widely utilized ultrasonography (USG) predictors in clinical practice,
owing to their simplicity and excellent performance. According to the original and
subsequent validation studies by the IOTA group, these rules could be applied in 76
to 78% of adnexal masses with a sensitivity and specificity of 92 to 94% and 91 to
95%, respectively.[2]
[3] However, IOTA-SR is limited by being inconclusive in a fair proportion of cases,
and these cases need further evaluation.[2]
[3]
The recently introduced three-dimensional (3D) power Doppler ultrasound offers the
potential for volume measurements and quantification of echogenicity and blood flow
in the whole target tissue, in contrast to the two-dimensional (2D) USG, which can
assess the vascularization in only one subjectively chosen 2D plane.[4] Additionally, it is highly reproducible between sonographers.[5] The role of 3D USG in gynecologic oncology has yet to be proven and needs to be
explored further. It is available only sometimes in daily clinical practice. Before
introducing this technique in routine use, it is crucial to know whether 3D USG has
added value for detecting malignancy over the widely practiced 2D ultrasound-based
simple rules. We conducted the present study to evaluate and compare the diagnostic
accuracy of 2D USG-based IOTA-SR and 3D USG for preoperative characterization of adnexal
masses.
Methods
A prospective observational study was conducted in the Department of Obstetrics and
Gynaecology and Department of Radiodiagnosis of All India Institute of Medical Sciences,
New Delhi, India, from January 2019 to August 2020. Women older than 18 years with
adnexal mass planned for surgical management were recruited in the study after signing
an informed consent and institutional ethical clearance (IECPG-478/29.11.2017). We
chose the mass with a more complex morphology or the larger size in patients with
bilateral masses. Pregnant women with an adnexal mass, patients who failed to undergo
surgery within 30 days of ultrasound examination, neoadjuvant chemotherapy recipients,
and patients with histopathologically confirmed diagnosis before surgery were excluded
from the study. Baseline characteristics and tumor marker values were recorded. Two
investigators conducted 2D and 3D USG with power Doppler examinations for all the
patients. For USG examination, both transvaginal sonography (TVS) and transabdominal
approaches (TAS) were used. TAS was used to examine large masses that could not be
visualized in their entirety using a transvaginal probe. Expert gynecologists performed
all the examinations using the Voluson E8 USG machine (4–8 MHz).
2D USG parameters noted were lesion diameter, septations, solid areas, acoustic shadow,
presence of and the number of papillary projections, wall irregularities, free fluid,
and echogenicity of the adnexal mass. After 2D grayscale ultrasound, tumor vascularity
was assessed using resistance index (RI), pulsatility index (PI), and peak systolic
velocity (PSV) by color Doppler examination. IOTA-SR using B and M features were applied,
and a presumptive diagnosis of benign, malignant, or inconclusive was made. The mass
was classified as malignant if one or more M features were present without B features.
The mass was classified as benign if one or more B features were present without M
features. If both B and M features could be applied or none of them were present,
the mass was classified as inconclusive.[1]
3D USG and power Doppler examination were then performed by another investigator blinded
to the 2D USG findings. The ultrasound examination was conducted using wall motion
filter low 2, at 0.8-kHz pulse repetition frequency (PRF), gain 0.8, with high quality
and line density of 8. 3D volume box was applied to cover the entire extent of the
mass if possible. The contour of the mass was outlined by selecting the manual mode
and after repeatedly rotating its image six times by 30 degrees, six tracings were
obtained to complete a 360-degree rotation. Additional information on cyst volume,
mean gray index (MGI), vascularization index (VI), flow index (FI), and vascularization
flow index (VFI) was obtained to characterize benign and malignant masses.[5] The virtual organ computer-aided analysis (VOCAL) imaging software installed in
the USG machine automatically calculated these indices ([Figs 1] and [2]). The power Doppler images were acquired at the following setting: maximum radiant
high-definition (HD) flow, with power doppler map 5, low wall motion filter, mid-frequency,
and PRF of 0.8kHz, which were preset in the GE Voluson by the engineer to optimize
image quality.
Fig. 1 (A) Two-dimensional grayscale image of an adnexal mass showing predominantly solid component
with few cystic areas. (B) Three-dimensional power Doppler volume calculation of the adnexal mass using manual
sampling. (C) Calculation of power Doppler indices using VOCAL software with histogram function.
Three-dimensional and power Doppler indices were cyst volume of 397.60 cm3, mean gray index (MGI) of 43.0, vascularization index (VI) of 33.0, flow index (FI)
of 51.06, and vascularization flow index (VFI) of 20.25, which were also suggestive
of a malignant nature. (D) Intraoperative picture of the adnexal mass. The final histopathology report was
suggestive of high-grade serous carcinoma.
Fig. 2 (A) Two-dimensional grayscale image showing a multilocular cystic adnexal mass. (B) Three-dimensional power Doppler volume calculation of an adnexal mass using manual
sampling method. (C) Calculation of power Doppler indices using VOCAL software with histogram function.
Three-dimensional and power Doppler indices were cyst volume of 460.21 cm3, mean gray index (MGI) of 24.3, vascularization index (VI) of 6.2, flow index (FI)
of 20.2, and vascularization flow index (VFI) of 1.2, suggestive of the benign nature
of the mass. (D) Intraoperative picture of the adnexal mass. The final histopathology report was
suggestive of serous cystadenoma.
The mean value of the grayscale voxels was calculated as the MGI (scale: 0–100) using
the “histogram” feature of the 3D View Function. MGI near zero refers to a sonolucent
mass; higher values represent increased echogenicity.
Surgery was performed by laparoscopy or laparotomy, as per the hospital protocol.
The excised mass underwent histopathological evaluation. According to the final histopathology
report, tumors were classified into benign and malignant. For analysis, tumors with
borderline histology were included in the malignant group.
2D USG-based IOTA-SR and 3D USG diagnostic accuracy were calculated using sensitivity,
specificity, and positive predictive value (PPV)/negative predictive value (NPV).
The study was powered to detect a 7% difference in sensitivity with 80% power and
95% confidence interval (95% CI; 5% level of two-sided significance). To assess the
significance of individual USG features for identifying malignancy, logistic regression
analysis was carried out by taking 2D and 3D ultrasound parameters separately. The
receiver operating characteristic (ROC) curve was plotted for significant variables,
and the areas under the curve (AUCs) with cut-off values using the Youden index were
calculated. Sensitivity and specificity for 2D USG-based simple rules and 3D ultrasound
were calculated. A two-sided probability of p < 0.05 was considered significant for all the statistical tests. The data analyses
were carried out using STATA version 12.0 software.
Results
Eighty-four patients underwent surgical management during the study period, and histologically
41 (48.8%) of these were classified as benign and 43(51.1%) as malignant. The mean
age (40.41 ± 16.83 vs. 46.05 ± 14.83 years; p = 0.103), body mass index (BMI; 23.97 ± 3.33 vs. 24.84 ± 2.53; p = 0.17), and mean parity (1.95 ± 1.92 vs. 1.79 ± 1.32; p = 0.65) of benign and malignant masses were comparable. The CA-125 concentration
(U/mL) was significantly higher in malignant masses than in the benign group (131.44 ± 374.88
vs. 763.19 ± 933.55; p < 0.001).
2D USG parameters are depicted in [Table 1]. Solid areas, random echogenicity, free fluid, and cyst wall irregularities were
significantly more common in the malignant masses. When the IOTA-SR were applied,
the masses were categorized as benign (39/74), malignant (35/74), and inconclusive
(10/84). When the diagnostic accuracy of individual rules was assessed, it was observed
that the B1 feature has the highest PPV of 93.3% for predicting the benign nature
of adnexal masses. The M5 feature had the highest PPV of 100%, followed by M3 (91.67%)
and M1 (91.30%), for predicting the malignant nature of the mass ([Table 2]). These observations were compared with the final histopathology report, and the
sensitivity, specificity, PPV, and NPV of IOTA-SR in conclusive cases were observed
as 83.78% (95% CI: 67.99–93.81%), 89.19% (95% CI: 74.58–96.97%), 88.6% (95% CI: 75.2–95.2),
and 84.62% (95% CI: 72.51–92.02), respectively.
Table 1
Two-dimensional (2D) and three-dimensional (3D) ultrasound parameters of benign and
malignant adnexal masses
|
Feature
|
Benign (N = 39)
|
Malignant (N = 35)
|
p Value
|
|
2D gray scale parameters
|
|
Mean cyst diameter (cm)
Mean ± SD (range)
|
10.4 ± 5.2 (2.3–29.3)
|
12.4 ± 5.3 (3.7–27.7)
|
0.085
|
|
Solid areas
|
29.3% (12/41)
|
90.7% (39/43)
|
<0.001
|
|
Septations
|
56.1% (23/41)
|
60.4% (26/43)
|
0.68
|
|
Acoustic shadow
|
12.2% (5/41)
|
11.6% (5/43)
|
0.93
|
|
Free fluid
|
9.7% (4/41)
|
44.1% (19/43)
|
<0.001
|
|
Randomly echogenic
|
31.7% (13/41)
|
79.07% (20.9%)
|
<0.001
|
|
Cyst wall irregularities
|
14.6% (6/41)
|
34.8% (15/43)
|
<0.001
|
|
2D Doppler parameters
|
|
PI (mean + SD)
|
1.02 ± 0.93
|
0.83 ± 0.63
|
0.315
|
|
RI (mean + SD)
|
0.75 ± 0.24
|
0.47 ± 0.24
|
0.001
|
|
PSV (cm/s)
|
27.25 ± 13.07
|
25.24 ± 11.36
|
0.51
|
|
Vessel characteristics
|
No flow
|
17 (41.46%)
|
3 (6.98%)
|
<0.001
|
|
Suspect
|
3 (7.31%)
|
23 (53.48%)
|
|
Nonsuspect
|
21 (51.23%)
|
17 (39.54%)
|
|
3D power Doppler parameters
|
|
Cyst volume (cm3)
|
|
462.71 ± 738.50
|
699.33 ± 799.16
|
0.163
|
|
MGI
|
|
22.29 ± 10.56
|
34.28 ± 8.67
|
<0.001
|
|
VI
|
|
5.08 ± 4.50
|
16.70 ± 10.40
|
<0.001
|
|
FI
|
|
22.96 ± 9.77
|
34.97 ± 7.15
|
<0.001
|
|
VFI
|
|
1.73 ± 1.59
|
6.32 ± 4.27
|
<0.001
|
Abbreviations: FI, flow index; MGI, mean gray index; PI, pulsatility index; PSV, peak
systolic velocity; RI, resistance index; VFI, vascularization flow index; VI, vascularization
index.
Table 2
Diagnostic accuracy of individual B and M rules
|
Sensitivity (%)
|
Specificity (%)
|
PPV (%)
|
NPV (%)
|
|
Diagnostic accuracy of B rules for predicting benign nature of adnexal mass
|
|
B1
|
34.15
|
97.67
|
93.33
|
60.87
|
|
B2
|
4.88
|
97.67
|
66.67
|
51.85
|
|
B3
|
12.20
|
88.37
|
50.00
|
51.35
|
|
B4
|
19.51
|
95.35
|
80.00
|
55.41
|
|
B5
|
41.46
|
93.02
|
85.00
|
62.50
|
|
Diagnostic accuracy of M rules for predicting malignant nature of adnexal mass
|
|
M1
|
48.84
|
95.12
|
91.30
|
63.93
|
|
M2
|
44.19
|
90.24
|
82.61
|
60.66
|
|
M3
|
25.58
|
97.56
|
91.67
|
55.56
|
|
M4
|
44.19
|
95.12
|
90.48
|
61.90
|
|
M5
|
32.56
|
100.00
|
100.00
|
58.57
|
Abbreviations: NPV, negative predictive value; PPV, positive predictive value.
Among the 3D power Doppler ultrasound parameters, MGI, VI, FI, and VFI were found
to be significantly associated with increased risk of malignancy using multivariate
logistic regression analysis (p = 0.036, 0.028, 0.032, and 0.019, respectively; [Table 1]). There are no established cut-off values for 3D ultrasound and power Doppler indices;
hence, the cut-off values for each were calculated using the ROC curve ([Fig. 3]). Among all the power Doppler indices, the VI showed the highest sensitivity of
83% for a cut-off value of 8.4 ([Table 3]). Overall, the sensitivity and specificity of 3D USG and power Doppler were 84 and
88%, respectively, with an AUC of 0.96 (95% CI: 0.92–0.99) and was comparable with
IOTA-SR.
Fig. 3 Receiver operating characteristic (ROC) curve for three-dimensional (3D) ultrasonography
(USG) parameters. (A) Vascularization index (cut-off: 8.4; area under the curve [AUC]: 0.89). (B) Mean gray index (cut-off: 28.2; AUC: 0.81). (C) Flow index (cut-off: 29.6; AUC: 0.86). (D) Vascularization flow index (cut-off: 3.1; AUC: 0.89).
Table 3
Sensitivity, specificity of estimated cut-off values for three-dimensional ultrasound
and power Doppler parameters
|
Variable
|
Cut-off value
|
Sensitivity
|
Specificity
|
AUC
|
95% CI
|
|
MGI
|
28.2
|
74%
|
73%
|
0.81
|
0.71–0.89
|
|
VI
|
8.4
|
83%
|
83%
|
0.89
|
0.82–0.96
|
|
FI
|
29.6
|
77%
|
76%
|
0.86
|
0.78–0.94
|
|
VFI
|
3.1
|
79%
|
78%
|
0.89
|
0.83–0.96
|
Abbreviations: AUC, area under the curve; CI, confidence interval; FI, flow index;
MGI, mean gray index; VFI, vascularization flow index; VI, vascularization index.
Inconclusive findings were observed when IOTA-SR were applied in 11.9% (10/84) cases.
Out of these 10 masses, both B and M features were present in 8 masses, and 2 had
none of the B and M features. Six were malignant, and four were benign, per the final
histopathology report. The histological diagnosis of 10 inconclusive cases is shown
in [Table 4]. These masses were irregular solid masses with the presence of acoustic shadow (B3M1),
multilocular irregular masses more than 10 cm with acoustic shadow (B3M4), multilocular
masses less than 10 cm with the presence of ascites (B4M2), and other combinations.
Of these 10 inconclusive cases, 4 were malignant, 2 were borderline tumors, 2 were
tuberculosis, and 1 each was endometrioma and serous cystadenoma. These masses were
triaged using 3D ultrasound with power Doppler. 3D parameters could correctly differentiate
40% (4/10) of SR inconclusive masses (a malignant disease in three and a benign disease
in one patient; [Table 4]).
Table 4
Three-dimensional characteristics of inconclusive cases
|
Sl. no.
|
|
Histopathology
|
3D USG characteristics
|
|
IOTA-SR characteristics
|
Cyst volume (cm3)
|
MGI (estimated cut-off: 28.2)
|
VI (estimated cut-off: 8.4)
|
FI (estimated cut-off: 29.6)
|
VFI (estimated cut-off: 3.1)
|
|
1
|
B4M2
|
High grade serous cancer (HGSC)
|
118.53
|
41.41
|
10.66
|
36.55
|
3.9
|
|
2
|
B3M1
|
Endometrioid adenocarcinoma
|
767.0
|
23.17
|
6.80
|
36.96
|
2.54
|
|
3
|
B1M2
|
Tuberculosis
|
239.00
|
30.38
|
10.40
|
29.22
|
3.05
|
|
4
|
No feature
|
Tuberculosis
|
100.45
|
28.34
|
7.60
|
46.33
|
3.50
|
|
5
|
B3M1M2
|
Endometrioma
|
97.81
|
13.00
|
16.24
|
25.00
|
4.10
|
|
6
|
B3M4
|
Dysgerminoma
|
831.00
|
47.20
|
10.31
|
34.34
|
3.54
|
|
7
|
B2M5
|
Endometrioid adenocarcinoma
|
337.02
|
30.29
|
15.83
|
39.54
|
6.26
|
|
8
|
No feature
|
Borderline mucinous tumor
|
691.26
|
32.15
|
16.91
|
29.11
|
4.92
|
|
9
|
B1B5M3
|
Borderline serous tumor
|
441.77
|
26.00
|
2.50
|
20.00
|
0.90
|
|
10
|
B5M3
|
Serous cystadenoma
|
666.32
|
0.00
|
0.00
|
0.00
|
0.00
|
Abbreviations: 3D, three-dimensional; FI, flow index; OITA-SR, International Ovarian
Tumuor Analysis Simple Rules; MGI, mean gray index; USG, ultrasonography; VFI, vascularization
flow index; VI, vascularization index.
Discussion
Adnexal masses are frequently encountered in clinical practice. The accurate preoperative
assessment facilitates timely referral, route, and extent of surgery. Clinical and
laboratory parameters like age, menopausal state, personal or family history of breast
or ovarian cancer, CA-125 level, and imaging are used to differentiate benign and
malignant masses. Various ultrasound models are available for preoperative assessment
of adnexal masses.
Among the USG-based reporting systems, IOTA-SR are commonly utilized because of their
simplicity and good diagnostic accuracy. Several investigators have evaluated the
diagnostic performance of IOTA-SR. In a systemic review, pooled sensitivity and specificity
of IOTA-SR were 0.93 (95% CI: 0.89–0.95) and 0.81, respectively.[6] Similar findings were observed by Meys et al, with a pooled sensitivity and specificity
of 0.93 (95% CI: 0.91–0.95) and 0.80 (95% CI: 0.77–0.82), respectively.[7] In the present study, sensitivity and specificity of simple rules were 83.78% (95%
CI: 67.99–93.81) and 89.19% (95% CI: 74.58–96.97), respectively, which were similar
to the results of a study reported by Auekitrungrueng et al, who reported sensitivity
and specificity of 83.8% (95% CI: 77.1–90.40) and 92.0 (95% CI: 88.8–95.2), respectively.[8] However, other authors have reported higher sensitivity to simple rules than our
observations.[9]
[10] The marginal variations reported in the diagnostic performance could be because
of the heterogeneous nature, variations in the prevalence of malignant masses, and
different levels of examiners conducting the test across these investigations.
The proportion of conclusive results using simple rules varied from 76 to 92% in different
studies.[1]
[3]
[8]
[9]
[11]
[12]
[13] The interpretation of simple rules is affected by the experience of the examiners.
In a study by Knafel et al, the proportion of conclusive results was 82.4% (lower
than our results) when level 1 examiners (with less experience) performed the examination;
this increased to 91.2% with level 2 examiners.[9] Most differences resulted from evaluating acoustic shadow, motion artefacts, and
subjective interpretation of color flow in adnexal masses. In our study, all the examinations
were done by experienced investigators, which explains the relatively higher proportion
of conclusive cases.
3D ultrasound allows visualization of adnexal masses in multiple planes, thus better
characterizing an adnexal mass. With surface rendering mode, the surface of an adnexal
mass can be visualized more precisely, and it also allows the 3D reconstruction of
vessels. Other features like inversion mode allow better visualization of areas of
cystic contents. With the addition of 3D power Doppler, the vascular architecture
can be visualized more clearly. In 3D Doppler, vascular indices (VI, FI, and VFI)
are calculated, which provides a more objective assessment of vascularity and would
probably decrease the subjective variation in the assessment of vascularity. However,
3D ultrasound is still under evaluation, and its use is limited by cost and availability.
In the current study, sensitivity and specificity for 3D USG for discriminating adnexal
masses were calculated as 84 and 88%, respectively, with AUC of 0.96 (95% CI: 0.92–0.99).
A study by Perez-Medina et al found a sensitivity and specificity of 84.6% and 81.9%,
respectively.[14] Huchon et al found a sensitivity and specificity of 82 and 90%, respectively, and
Alcazar et al reported a sensitivity and specificity of 97.8 and 79.2%, respectively.[15]
[16] A summary of available studies conducted to evaluate the efficacy of 3D USG is depicted
in [Table 5].[4]
[14]
[15]
[17]
[18]
[20]
Table 5
Review of studies evaluating diagnostic performance of 3D USG
|
Study
|
Mean VI
|
Mean FI
|
Mean VFI
|
|
Abbas et al[17]
|
10.98 ± 9.17 vs. 16.36 ± 15.18; p < 0.05
|
20.15 vs. 20.16; p > 0.05
|
2.13 ± 2.01 vs. 3.91 ± 3.83; p < 0.01
|
|
Perez et al[14]
|
5.38 ± 6.61 vs. 6.29 ± 5.77; p = 0.53
|
29.63 ± 10.29 vs. 33.81 ± 10.39; p = 0.15
|
1.68 ± 2.10 vs. 2.37 ± 2.72; p = 0.24
|
|
Huchon et al[15]
|
7.2 ± 8.0 vs. 35.5 ± 20.8; p < 0.0001
|
37.0 ± 11.5 vs 48.2 ± 11.0; p = 0.003
|
2.9 ± 3.6 vs. 17.6 ± 12.5; p < 0.0001
|
|
Ohel et al[4]
|
6.5 ± 4.2 vs. 6.2 ± 4.6
|
41.6 ± 9.8 vs. 36.0 ± 8.6
|
|
|
Jokubkiene et al[18]
|
5.1 (0.03–60.53) vs. 35.6 (4.73–78.61); p < 0.001
|
|
|
|
Wilson et al[20]
|
1.3 ± 1.6 vs. 4.7 ± 3.9; p < 0.01
|
|
|
|
Present study
|
5.08 ± 4.50 vs. 16.70 ± 10.40; p < 0.001
|
22.96 ± 9.77 vs. 34.97 ± 7.15; p < 0.001
|
1.73 ± 1.59 vs. 6.32 ± 4.27; p < 0.001
|
Abbreviations: 3D, three-dimensional; FI, flow index; VFI, vascularization flow index;
USG, ultrasonography; VI, vascularization index.
Our study has demonstrated numerically increased cyst volume, significantly increased
MGI, and raised vascular indices in malignant masses ([Table 1]). Malignant masses are highly vascular due to neovascularization and the vascularization
index reflects the density of blood vessels in an adnexal mass. The results of our
study corroborated with the existing data, which also demonstrated a higher value
of VI in malignant masses (16.70 ± 10.40 vs. 5.08 ± 4.50). A wide variation is observed
in absolute VI values in different studies. However, all studies have consistently
demonstrated higher values of VI in malignant masses, except the studies by Perez-Medina
et al and Ohel et al, where no significant difference was demonstrated between the
two benign and malignant masses, which could be attributed to the small study sample
size.[4]
[14] The observed variation in the value of vascular indices may be due to a difference
in technique and lack of standardization. FI is the sum of weighted color voxels divided
by the total number of color voxels. It reflects the number of blood corpuscles in
the vessels of the volume examined. Additionally, the chaotic architecture of vessels
and complex branching patterns in adnexal masses were noted. The results of our study
are consistent with the study by Huchon et al, which also showed higher values of
FI in malignant masses.[15] However, other studies did not find any significant differences ([Table 5]). Hence, FI's role in discriminating the adnexal masses must be explored further.
VFI is the total number of weighted color voxels divided by the total number in the
region of interest. It reflects the density of blood within the region and the number
of corpuscles in the vessels in the volume. Like our observations, a higher value
of VFI in malignant masses was observed in the published literature.
Several authors have investigated the cut-off values of Doppler indices, and different
cut-off values with variable accuracy have been shown in the literature. A cut-off
value of 3.1 for VFI has demonstrated sensitivity and specificity of 79 and 78%, respectively,
for the detection of malignancy. Geomini et al calculated the VFI cut-off value of
2.0 to predict the risk of malignancy with an odds ratio of 0.74 (0.45–1.23) for VFI
<2 and 0.92 (0.74–1.25) for VFI >2.[19] These authors also calculated the cut-off value of FI as 30 to predict the risk
of malignancy with an odds ratio of 0.83 (0.74–0.94) (95% CI) for FI <30 and 1.07
(0.99–1.15) for FI >30. Our study's cut-off value of 29.6 for FI demonstrated a sensitivity
and specificity of 77 and 76%, respectively. Wilson et al suggested a cut-off value
2.3 for VI with a sensitivity and specificity of 75% and 90%, respectively.[20] In the present study, the cut-off value of 8.4 for VI demonstrated both sensitivity
and specificity of 83%. Despite similar observations, the absolute value of VI was
different in their study than ours, and this could be explained by the fact that our
study used power processing and frequency-based color Doppler.
Hence, it is evident that there has yet to be a clear cut-off value established so
far for the power Doppler indices to differentiate between benign and malignant masses
due to variations in tissue attenuation and machine settings, such as gain and pulse
repetition by different examiners. There is insufficient evidence to support the application
of this technology in clinical practice.
Both modalities (IOTA-SR and 3D ultrasound) have demonstrated good diagnostic performance
in the studies. The significant benefits of IOTA-SR over 3D ultrasound are that they
are easy to learn, simple to use with a short learning curve, and can be used by nonexpert
examiners.[8]
[20] The main concern with IOTA-SR remains the proportion of inconclusive cases. Several
strategies like three-step assessment, logistic regression models, Assessment of Different
NEoplasias in the adneXa (ADNEX model), and expert USG assessment have been investigated.[21]
[22] We also evaluated the role of 3D ultrasound with power Doppler for evaluating the
IOTA inconclusive cases. 3D USG could correctly classify three malignant masses and
one benign mass out of 10 inconclusive cases. So, 3D ultrasound with power Doppler
could be used as a second-stage test to evaluate the inconclusive masses. Nevertheless,
this needs further confirmation in large-scale prospective studies.
The study's strength was that the comparison of both techniques was performed using
the same ultrasound machine and settings, which allowed an ideal comparison of both
techniques. The limitations were the small sample size and the ultrasound was performed
only for patients planned for surgery.
Conclusion
IOTA-SR and 3D USG have similar diagnostic performances in discriminating benign and
malignant adnexal masses. 3D USG does not provide any added advantage over IOTA-SR.
However, the potential use of 3D ultrasound as a second-stage test in inconclusive
masses should be further evaluated. More large-scale studies are required to develop
a standardized technique and cut-off values for power Doppler parameters before its
implementation in routine clinical practice.
