Suh Burgmann2021

You might also like

Download as pdf or txt
Download as pdf or txt
You are on page 1of 8

Original Research ajog.

org

GYNECOLOGY
Ultrasound characteristics of early-stage high-grade
serous ovarian cancer
Elizabeth Suh-Burgmann, MD; Natasha Brasic, MD; Priyanka Jha, MBBS; Yun-Yi Hung, PhD; Ruth B. Goldstein, MD

BACKGROUND: Survival from ovarian cancer is strongly dependent on RESULTS: Among 111 women identified, 4 had bilateral ovarian
the stage at diagnosis. Therefore, when confronted with a woman with an involvement, for a total of 115 adnexal masses characterized by
isolated adnexal mass, clinicians worry about missing the opportunity to ultrasound examination. The mean age at diagnosis was 61.8 years
detect cancer at an early stage. High-grade serous ovarian cancers ac- (range, 42e91 years). The median mass size was 9.6 cm (range,
count for 80% of ovarian cancer deaths, largely because of their tendency 2.2e23.6 cm) with 87% of cases having a mass size of 5 cm. A
to be diagnosed at a late stage. Among adnexal masses, large size and the mixed cystic and solid appearance was most common (77.4%), but a
presence of solid areas on ultrasound examination have been found to be completely solid appearance was more frequently seen for tumors of
associated with cancer, but it is unclear whether these characteristics <5 cm compared with larger tumors (26.7% vs 13.0%). Solid
identify early-stage cases. components other than septations were seen in 97.4% of cases. The
OBJECTIVE: This study aimed to evaluate the ultrasound findings characteristics of stage I and II cases were similar other than ascites,
associated with clinically detected early-stage high-grade serous ovarian which was more commonly seen in stage II cases (18.0% vs 3.1%,
cancer. respectively). Interobserver concordance was high for size and vol-
STUDY DESIGN: This was a retrospective cohort study of women ume measurements (correlation coefficients, 0.96e0.99), with
diagnosed with stage I or II high-grade serous ovarian or fallopian tube moderate agreement observed across the other ultrasound charac-
cancer measuring at least 1 cm at pathology from 2007 to 2017. Pre- teristics (Fleiss kappa, 0.45e0.58).
operative ultrasound examinations were independently reviewed by 3 CONCLUSION: In this community-based cohort, early-stage high-
radiologists. Adnexal masses were scored for size and volume; overall grade serous cancers rarely presented as masses of <5 cm or masses
appearance; presence, thickness, and vascularity of septations; without solid components other than septations. Our findings provide
morphology and vascularity of other solid components; and degree of additional support for the observation of small masses without solid areas
ascites. Characteristics were compared between masses of <5 cm and on ultrasound examination.
larger masses and between stage I and stage II cases. Interobserver
variability was assessed. Key words: cancer, cyst, early detection, ovary, ultrasound

Introduction screening has not resulted in improve- concern for cancer even when worri-
Ovarian cancer is the most lethal of gy- ment in ovarian cancer mortality.4,5 some characteristics are absent.
necologic cancers, responsible for Therefore, when confronted by a Therefore, we analyzed the ultrasound
approximately 22,000 deaths annually.1 woman with an adnexal mass, clinicians characteristics of early-stage high-grade
Stage is strongly associated with prog- often worry about missing the oppor- serous cancers among women diagnosed
nosis, with 90% survival for women tunity to detect cancer at an early stage. over a 10-year period in a large
with stage I cancer compared with 15% The ultrasound characteristics of large community-based setting.
to 20% survival for women with stage III mass size and the presence of solid areas
or IV cancer.2 Of the various types of on ultrasound examination have been Methods
ovarian cancer, high-grade serous can- shown in several studies to be strongly This was a retrospective cohort study of
cers account for approximately 80% of associated with malignancy.6e12 How- women diagnosed with stage I or II high-
deaths, largely because of their tendency ever, it is unclear to what degree these grade serous ovarian cancer from 2007
to be detected at a late stage.3 The failure ultrasound characteristics help identify to 2017. Participants were members of
to detect high-grade serous cancers at the early-stage cancers, as most cancers Kaiser Permanente Northern California
early stage is a major reason why identified in studies, particularly high- (KPNC), a closed, integrated healthcare
grade serous cancers, were late stage. As delivery system, including 21 hospitals
early stage cases are relatively rare, no that provide care for more than 4.2
Cite this article as: Suh-Burgmann E, Brasic N, Jha P, studies have specifically evaluated the million members whose racial and
et al. Ultrasound characteristics of early-stage high-grade ultrasound characteristics of early-stage ethnic diversity mirrors that of the
serous ovarian cancer. Am J Obstet Gynecol ovarian high-grade serous cancers, communities served. The study period
2021;XX:x.exex.ex. despite the importance of their recog- was selected as the period for which
0002-9378/$36.00 nition. The resulting uncertainty complete tumor registry and electronic
ª 2021 Elsevier Inc. All rights reserved. regarding the ultrasound appearance of medical record data were available.
https://doi.org/10.1016/j.ajog.2021.04.262
early-stage cancers contributes to Approval for the study was obtained

MONTH 2021 American Journal of Obstetrics & Gynecology 1.e1


Original Research GYNECOLOGY ajog.org

review; and women with tumors


AJOG at a Glance measuring <1 cm at pathology,
Why was this study conducted? including serous tubal carcinoma in situ
Although high-grade serous cancers account for 80% of ovarian cancer deaths, no and cases that were associated with a
study has evaluated the ultrasound characteristics of these cancers at an early concurrent benign adnexal neoplasm.
stage. Using electronic databases linked to the
medical record, all pelvic ultrasound
Key findings examinations performed within 1 year
Over 87% of early-stage high-grade cancers presented with an adnexal mass of >5 before cancer diagnosis were identified.
cm, with a median size of 9.6 cm. In 97% of cases, the mass demonstrated solid If women had more than 1 ultrasound
components other than septations. during this period, the first study was
used for analysis to characterize the mass
What does this add to what is known? at the earliest point of detection.
Our findings have indicated that although small adnexal masses are commonly The indication for ultrasound exami-
detected by ultrasound examination, they are an uncommon presentation of nation was determined by manual re-
early-stage high-grade serous ovarian cancer. The findings provide additional view of the “Indication” section of the
support for the observation of small masses without solid components. ultrasound report. In cases where the
stated indication was to follow up a
finding from another imaging study,
from the KPNC Institutional Review cancer) were identified from the in- such as computed tomography (CT), the
Board with waivers of informed consent stitution’s cancer registry, with manual “Indication” section for the other imag-
because of the retrospective nature of the medical record review to confirm diag- ing study was also reviewed to determine
study. nosis, stage, and histology. We excluded if it was potentially related to the mass.
Women with stage I or II high-grade women who did not undergo surgical For example, if the indication for ultra-
serous ovarian or fallopian tube cancer staging; women with borderline, low- sound examination was to follow up a
(henceforth referred together as ovarian grade, or nonserous histology on CT finding and the CT was done for
abdominal pain, the indication for the
FIGURE 1 ultrasound examination was considered
Flowchart to be pain, whereas if the CT was done
for a problem with no potential rela-
Women with stage I-II ovarian or fallopian tionship to the mass, such as surveillance
tube cancer diagnosed 2007-2017 with for abdominal aneurysm, the indication
serous histology was considered to be an incidental
we
N=257 women finding.
Ultrasound imaging files were dei-
dentified of all protected health infor-
Had pelvic ultrasound within 1 year prior to mation and independently reviewed by 3
Excluded: N=84, no ultrasound done within
diagnosis
1 year of diagnosis
radiologists (N.B., P.J., and R.B.G.), all of
N=173 women whom specialize in ultrasound and
women’s imaging. Each radiologist
independently assessed each ultrasound
Excluded: N=39 (not mutually exclusive)
imaging for technical adequacy to eval-
Low grade or other histology N=20 uate the adnexa, based on the quality of
Review of pathology confirmed stage I-II,
high-grade with
Staging not done or incomplete N=2 the available images. Studies deemed
Figure 2 displays the> mass
1 cm tumor
g at pathology
Not stage I-II N=6
inadequate for scoring because of tech-
N=134 women Tumor < 1 cm N=11
Concurrent benign neoplasm N=1 nical factors by at least 2 radiologists
were subsequently excluded from the
analysis.
Adnexal masses were characterized in
Ultrasound review by 3 radiologists
terms of (1) largest size dimension (in
Excluded: N=23, inadequate study quality
millimeters); (2) size dimensions in 3
N=111 women with 115 masses (4 cases
with bilateral cancer)
planes (in millimeters); (3) overall
appearance (mixed cystic and solid,
Flowchart of the study cohort. solid, multilocular, nonsimple unilocu-
Suh-Burgmann et al. Ultrasound appearance of early-stage ovarian cancer. Am J Obstet Gynecol 2021.
lar, or simple unilocular); (4) septations
(none, single or multiple, thick [ 3

1.e2 American Journal of Obstetrics & Gynecology MONTH 2021


ajog.org GYNECOLOGY Original Research

mm] or thin); (5) vascularity of septa-


TABLE 1
tions (NA [not applicable], no, yes); (6)
Patient characteristics
presence of other solid components (yes
or no); (7) morphology of solid areas Characteristic Value
(NA, smooth, entire mass is solid, Age at ultrasound (y)
irregular or papillary); (8) vascular flow
MeanSD 62.0 (11.5)
in solid component (NA, no, yes); and
(9) ascites (none or mild, moderate or Median (IQR) 61 (53e68)
severe). Normal postmenopausal ovaries Range, n (%) 42e91
were differentiated from small solid 18e39 0 (0)
masses based on irregularities of shape
40e49 14 (12.6)
or margins and presence of abnormal
internal flow. Septations were defined as 50e59 40 (36.0)
structures within cystic masses that tra- 60e69 31 (27.9)
verse from 1 wall to another and 70 26 (23.4)
considered thin if <3 mm in the greatest
width. If >1 cm bilateral ovarian tumor Race, n (%)
involvement was demonstrated on sur- White 68 (61.3)
gical pathology, both adnexa were scored Black 9 (8.1)
separately and treated as independent Hispanic 13 (11.7)
cases. Mass size was calculated as the
Asian or Pacific Islander 18 (16.2)
mean of the largest dimension as recor-
ded by 3 individual measurements. For Other 3 (2.7)
categorical features, agreement of at least Days from ultrasound to diagnosis
2 of the 3 radiologists was needed. If all 3 Median (IQR) 20 (10e36)
radiologists disagreed on how a feature
should be categorized, the study was re- Range 0e165
reviewed by all radiologists to arrive at a Primary location, n (%)
consensus. Ovary 98 (88.3)
The sonographic characteristics of
Fallopian tube 13 (11.7)
masses were compared for stage I vs II
cancer and for smaller vs larger masses. Stage, n (%)
We selected 5 cm as the size cutoff for I 63 (56.8)
comparison based on the view that II 48 (43.2)
masses of <5 cm are generally regarded as
BRCAþ or other deleterious mutation, n (%) 23 (20.7)
“small” from a clinical perspective. For
IQR, interquartile range; SD, standard deviation.
descriptive statistics, chi-square or the
Suh-Burgmann et al. Ultrasound appearance of early-stage ovarian cancer. Am J Obstet Gynecol 2021.
Fisher exact tests were used for categorical
variables, and unpaired t test was used for
continuous variables. Significance was identified. The inclusion and exclusion woman had surgery for large fibroids. Of
considered met at P<.05. Interobserver criteria are detailed in Figure 1. Among the 111 women included, 4 had bilateral
variability was assessed on the basis of those excluded were 11 women with adnexal cancer involvement for a total of
scores before re-review of discordant tubal carcinomas measuring <1 cm 115 abnormal adnexa scored. The most
scores. Correlation coefficients were (mean, 3.3 mm; range, 1e6 mm). In 7 of common indication for ultrasound ex-
determined for continuous variables, these cases, the occult tubal carcinoma amination was pain (33.3%), followed
such as mass size. For categorical vari- was found at the time of risk-reducing by mass on examination (23.4%), post-
ables, such as presence of ascites, the surgery for women known to be at menopausal bleeding (12.6%), and
Fleiss kappa was assessed. All statistical elevated risk because of a germline incidental finding on other imaging tests
analyses were performed using the Sta- BRCA mutation, with preoperative ul- (11.7%).
tistical Analysis System software (version trasound examination read as normal in The mean age at the time of surgery
9.4; SAS Institute Inc, Cary, NC). each case. For the remaining 4 women, was 61.8 years (median, 61 years; range,
occult tubal cancer was found inciden- 42e91 years). In all cases, the diagnosis
Results tally at the time of surgery that was of cancer was made on surgical pa-
After exclusions were applied, 111 performed for an unrelated abnormal thology, with a median time from index
women diagnosed with stage I or II high- ultrasound finding: 3 women had con- ultrasound examination to surgery of
grade serous cancer at surgery were current large benign ovarian cysts and 1 20 days (range, 0e165 days). Slightly

MONTH 2021 American Journal of Obstetrics & Gynecology 1.e3


Original Research GYNECOLOGY ajog.org

malignant.15 An association between


FIGURE 2
solid components on ultrasound exam-
Size distribution of masses
ination and risk of malignancy was
observed in the ultrasound screening
arm of the UK Collaborative Trial of
Ovarian Cancer Screening, which found
that masses without solid elements had
an absolute risk of 0.4% compared with
masses with solid elements that had an
absolute risk of 4.45%.11 In addition, our
results are generally consistent with the
ultrasound characteristics defined by the
International Ovarian Tumor Analysis
group’s 5 “Simple Rules” for identifying
ovarian malignancy: (1) irregular solid
tumor, (2) ascites, (3) at least 4 papillary
structures, (4) irregular multilocular-
solid tumor with the largest diameter
Associated with early-stage high-grade serous carcinoma of at least 10 cm, and (5) very high color
Suh-Burgmann et al. Ultrasound appearance of early-stage ovarian cancer. Am J Obstet Gynecol 2021.
content on color Doppler examination.8
Although the presence of solid compo-
nents on ultrasound examination has
more women were diagnosed with stage Interobserver concordance for size been found in screening trials and other
I cancer (56.5%) vs stage II cancer. At measurements was high with correlation studies to be associated with cancer,6e12
the time of data analysis, 23 women coefficients across the 3 radiologists of most high-grade serous cancers in those
(20.7%) had been found to carry a 0.98 to 0.99 for the largest size dimen- studies were late stage. Our study sug-
BRCA or other deleterious germline sion of the mass and 0.96 to 0.98 for gests that this association holds even for
mutation (Table 1). Among these volume as calculated by measurements early-stage cancers.
women, the average age at ultrasound in 3 planes. For categorical features, the
examination was slightly younger at 59 Fleiss kappa ranged from 0.45 to 0.58, Clinical implications
years. indicating moderate agreement among Our findings indicate that although
The median size of the associated the 3 radiologists across the other small adnexal masses are commonly
adnexal mass on ultrasound examina- characteristics. detected by ultrasound examination,
tion was 9.6 cm (range, 2.2e23.6 cm; they are an uncommon presentation of
interquartile range, 6.8e13.6 cm), with Comment early-stage high-grade serous ovarian
87% of tumors being at least 5 cm in size Principal findings cancer. We previously reported that
(Figure 2). Overall, a “mixed cystic and In this community-based cohort, we among 1363 women over the age of 50
solid” appearance was most common found that early-stage high-grade serous years with clinically detected complex
(77.4%), followed by “solid” appearance cancers rarely present as masses of <5 adnexal masses of <6 cm, the inci-
(14.8%). However, among tumors of <5 cm or as masses without solid areas other dence of cancer was 0.5%, and all
cm, 53.3% were “mixed cystic and solid” than septations on ultrasound cancers that were initially observed
in appearance, with 26.7% having a examination. demonstrated growth on the first
“solid” appearance compared with follow-up ultrasound examination
11.7% of larger tumors. There was no Results with no apparent negative impact on
case associated with a simple unilocular The finding that the median mass size in stage at diagnosis.16 Given that surgical
cyst. The presence of solid components these early-stage cases was 9.6 cm was morbidity tends to be higher for older
other than septations was the most somewhat surprising given that high- women, ultrasound observation may
consistent characteristic for both smaller grade serous cancers are thought to be be appropriate when balancing poten-
and larger tumors, seen in 97.4% of cases often detected at advanced stages tial benefits vs risks of surgical inter-
overall and demonstrating vascularity in because of their tendency to metastasize vention. The fact that nearly all
74.8% of cases (Table 2). Ascites was the early, before the primary tumor grows early-stage cancers demonstrated solid
only characteristic that differed by stage large enough to produce symptoms.13,14 areas other than septations on initial
with moderate-to-severe ascites seen in However, this finding is consistent with ultrasound examination provides
18% of stage II and 3.1% of stage I cases data from the Prostate Lung Colorectal additional reassurance and support for
(Table 3). Representative images are Ovarian cancer screening trial, which observation of small masses without
showing in Figure 3. found that masses of <5 cm were rarely solid components.

1.e4 American Journal of Obstetrics & Gynecology MONTH 2021


ajog.org GYNECOLOGY Original Research

TABLE 2
Ultrasound characteristics of adnexal masses associated with early-stage serous cancer by size
Variables Total (N¼115) Mass <5.0 cm (n¼15) Mass 5.0 cm (n¼100) P value
Appearance .019
Mixed cystic and solid 89 (77.4) 8 (53.3) 81 (81.0)
Multilocular 6 (5.2) 3 (20.0) 3 (3.0)
Solid 17 (14.8) 4 (26.7) 13 (13.0)
Nonsimple unilocular 3 (2.6) 0 3 (3.0)
Simple unilocular 0 0 0
Septation .289
Multiple 67 (58.3) 6 (40.0) 61 (61.0)
Single 9 (7.8) 2 (13.3) 7 (7.0)
None 39 (33.9) 7 (46.7) 32 (32.0)
Vascularity of septation .088
No 40 (34.8) 7 (46.7) 33 (33.0)
Yes 36 (31.3) 1 (6.7) 35 (35.0)
NA 39 (33.9) 7 (46.7) 32 (32.0)
Solid area other than septation .345
No 3 (2.6) 1 (6.7) 2 (2.0)
Yes 112 (97.4) 14 (93.3) 98 (98.0)
Vascular flow in solid component .466
No 26 (22.6) 3 (20.0) 23 (23.0)
Yes 86 (74.8) 11 (73.3) 75 (75.0)
NA 3 (2.6) 1 (6.7) 2 (2.0)
Morphology of the solid areas .032
Entire mass is solid 17 (14.8) 4 (26.7) 13 (13.0)
Irregular and/or papillary 85 (73.9) 7 (46.7) 78 (78.0)
Smooth 10 (8.7) 3 (20.0) 7 (7.0)
NA 3 (2.6) 1 (6.7) 2 (2.0)
Ascites 1.000
None or mild 104 (90.4) 14 (93.3) 90 (90.0)
Moderate or severe 11 (9.6) 1 (6.7) 10 (10.0)
Data are presented as number (percentage), unless otherwise indicated.
NA, not applicable.
Suh-Burgmann et al. Ultrasound appearance of early-stage ovarian cancer. Am J Obstet Gynecol 2021.

Although the optimal schedule of ul- likelihood of incidental findings. In our limited follow-up with repeat ultrasound
trasound examination follow-up has not setting, clinical management decisions examination at 6 weeks and 6 months,
been defined, studies indicate that many are supported by a standardized ultra- whereas those with intermediate risk
masses require several months to resolve; sound reporting system that stratifies have repeat ultrasound examination at 6
as such, a longer time interval has a masses as low, intermediate, or high risk weeks and 3, 6, 12, and 24 months. In-
higher likelihood of demonstrating res- based on ultrasound characteristics.9 In dividuals with masses with high-risk
olution.17 The observation that a mass is the absence of other clinical indicators of characteristics, which we define as solid
stable is also more meaningful the longer malignancy, such as elevated CA 125 areas larger than 1 cm and abnormal
the time interval. However, repeated ul- tumor markers, postmenopausal women vascular flow, are referred to gynecologic
trasound examinations may increase the with masses considered low risk undergo oncology.

MONTH 2021 American Journal of Obstetrics & Gynecology 1.e5


Original Research GYNECOLOGY ajog.org

TABLE 3
Ultrasound characteristics of adnexal masses associated with stage I vs stage II high-grade serous cancer
Variables Total (N¼115) Stage I (n¼65) Stage II (n¼50) P value
Age 1.000
MeanSD 61.811.4 61.811.2 61.811.9
Mass size .370
MeanSD 10.54.7 10.85.1 10.04.1
MinimumeMaximum 2.2e23.6 2.2e23.6 3.3e20.7
Median (IQR) 9.6 (6.8e13.6) 10.9 (6.9e14.5) 9.4 (6.8e13.2)
Race and ethnicity .416
White 71 (61.7) 41 (63.1) 30 (60.0)
Black 9 (7.8) 6 (9.2) 3 (6.0)
Hispanic non-Black 13 (11.3) 7 (10.8) 6 (12.0)
Asian or Pacific Islander 19 (16.5) 8 (12.3) 11 (22.0)
Other 3 (2.6) 3 (4.6) 0 (0)
Appearance .286
Mixed cystic and solid 89 (77.4) 53 (81.5) 36 (72.0)
Multilocular 6 (5.2) 4 (6.2) 2 (4.0)
Solid 17 (14.8) 6 (9.2) 11 (22.0)
Nonsimple unilocular 3 (2.6) 2 (3.1) 1 (2.0)
Simple unilocular 0 (0) 0 (0) 0 (0)
Septation .404
Multiple 67 (58.3) 37 (56.9) 30 (60.0)
Single 9 (7.8) 7 (10.8) 2 (4.0)
None 39 (33.9) 21 (32.3) 18 (36.0)
Vascularity of septation .915
No 40 (34.8) 23 (35.4) 17 (34.0)
Yes 36 (31.3) 21 (32.3) 15 (30.0)
NA 39 (33.9) 21 (32.3) 18 (36.0)
Other solid components .256
No 3 (2.6) 3 (4.6) 0 (0)
Yes 112 (97.4) 62 (95.4) 50 (100.0)
Vascular flow in solid components .193
No 26 (22.6) 17 (26.2) 9 (18.0)
Yes 86 (74.8) 45 (69.2) 41 (82.0)
NA 3 (2.6) 3 (4.6) 0 (0)
Morphology of the solid areas .123
Entire mass is solid 17 (14.8) 6 (9.2) 11 (22.0)
Irregular and/or papillary 85 (73.9) 51 (78.5) 34 (68.0)
Smooth 10 (8.7) 5 (7.7) 5 (10.0)
NA 3 (2.6) 3 (4.6) 0 (0)
Suh-Burgmann et al. Ultrasound appearance of early-stage ovarian cancer. Am J Obstet Gynecol 2021. (continued)

1.e6 American Journal of Obstetrics & Gynecology MONTH 2021


ajog.org GYNECOLOGY Original Research

TABLE 3
Ultrasound characteristics of adnexal masses associated with stage I vs stage II high-grade serous cancer (continued)
Variables Total (N¼115) Stage I (n¼65) Stage II (n¼50) P value
Ascites .010
None or mild 104 (90.4) 63 (96.9) 41 (82.0)
Moderate or severe 11 (9.6) 2 (3.1) 9 (18.0)
IQR, interquartile range; NA, not applicable; SD, standard deviation.
Suh-Burgmann et al. Ultrasound appearance of early-stage ovarian cancer. Am J Obstet Gynecol 2021.

In addition, observation is generally compared the clinical presentations of women to be diagnosed at advanced
reserved for asymptomatic women, women with early- and late-stage stages, our findings suggest that there
whereas most of the women with early- ovarian cancers. In that study, a may be a subset that grows confined to
stage cancer in this study were symp- palpable mass and postmenopausal the ovary for a prolonged period and is
tomatic with only 12% of cases being bleeding were the only 2 presenting therefore amenable to early-stage detec-
detected because of incidental findings symptoms more likely to be associated tion. The observation that a completely
on imaging. The 3 most common in- with early-stage vs late-stage ovarian solid appearance was more common
dications for ultrasound examination cancer diagnosis.18 among smaller than larger masses also
were abdominal pain, a palpable mass on suggest that the ultrasound appearance
examination, and postmenopausal Research implications of some cancers may evolve from a pre-
bleeding, which is consistent with a Although many high-grade serous can- dominantly solid appearance to a more
previously published study that cers metastasize early, leading 80% of cystic and solid appearance as the tumor
grows.

FIGURE 3 Strengths and limitations


Representative ultrasound images and characteristics The strengths of the study include the
community-based, well-characterized
and diverse nature of the cohort, which
increased the generalizability of our
findings to other community settings.
Despite the rarity of clinically detected
early-stage high-grade serous cancer, the
ability to draw from 10 years of experi-
ence in a large integrated health system
setting enabled the identification of these
women and their preceding ultrasound
studies. Ultrasound examinations were
independently scored by 3 radiologists,
reducing the potential impact of a single
individual’s style or preferences on ul-
trasound characterization. The limita-
tions of the study included its
retrospective nature and that the radi-
ologists were not blinded to the fact that
these were ultrasound examinations of
women who had been diagnosed with
cancer, which may have influenced their
ultrasound scoring. The quality of ul-
trasound examinations also varied,
limiting the ability to assess all charac-
teristics equally across patients. As our
study evaluated only high-grade serous
Suh-Burgmann et al. Ultrasound appearance of early-stage ovarian cancer. Am J Obstet Gynecol 2021.
cancers, our conclusions regarding the
safety of observation related only to the

MONTH 2021 American Journal of Obstetrics & Gynecology 1.e7


Original Research GYNECOLOGY ajog.org

risk of missing a high-grade serous can- Control and Prevention. 2020. Available at: 12. Buys SS, Partridge E, Black A, et al. Effect of
cer, not other ovarian cancer subtypes. www.cdc.gov/cancer/dataviz. Accessed screening on ovarian cancer mortality: the
January 10, 2021. Prostate, Lung, Colorectal and Ovarian (PLCO)
However, unlike high-grade serous can- 2. US Department of Health and Human Ser- Cancer Screening randomized controlled trial.
cer, the other common histologic types vices, National Institutes of Health, National JAMA 2011;305:2295–303.
of ovarian cancer (endometrioid, clear Cancer Institute. Cancer stat facts: ovary can- 13. Shih IeM, Kurman RJ. Ovarian tumorigen-
cell, mucinous) usually come to clinical cer. Available at: http://seer.cancer.gov/ esis: a proposed model based on morphological
attention at the early stage because of statfacts/html/ovary.html. Accessed January and molecular genetic analysis. Am J Pathol
10, 2021. 2004;164:1511–8.
their tendency to grow to a large size 3. Bowtell DD, Böhm S, Ahmed AA, et al. 14. Brown PO, Palmer C. The preclinical natural
before metastasizing.19 Finally, the study Rethinking ovarian cancer II: reducing mortality history of serous ovarian cancer: defining the
did not establish what the earliest from high-grade serous ovarian cancer. Nat Rev target for early detection. PLoS Med 2009;6:
detectable ultrasound abnormality of Cancer 2015;15:668–79. e1000114.
these tumors potentially could have 4. Henderson JT, Webber EM, Sawaya GF. 15. Partridge EE, Greenlee RT, Riley TL,
Screening for ovarian cancer: updated evidence et al. Assessing the risk of ovarian malig-
been but rather reports what it was in report and systematic review for the US Pre- nancy in asymptomatic women with
actual practice. Apart from the ventive Services Task Force. JAMA 2018;319: abnormal CA 125 and transvaginal ultra-
described exclusions, all clinically 595–606. sound scans in the prostate, lung, colorectal,
detected high-grade serous cancers and 5. United States Preventive Services Task and ovarian screening trial. Obstet Gynecol
all ultrasound examinations performed Force, Grossman DC, Curry SJ, et al. Screening 2013;121:25–31.
for ovarian cancer: US Preventive Services Task 16. Suh-Burgmann E, Hung YY, Kinney W.
within a year of diagnosis were included Force recommendation statement. JAMA Outcomes from ultrasound follow-up of small
in the study. Therefore, although it is 2018;319:588–94. complex adnexal masses in women over 50. Am
possible that these cancers could have 6. Levine D, Brown DL, Andreotti RF, et al. J Obstet Gynecol 2014;211:623.e1–7.
come to clinical attention when smaller, Management of asymptomatic ovarian and 17. Pavlik EJ, Ueland FR, Miller RW, et al.
they simply did not, suggesting that the other adnexal cysts imaged at US: Society of Frequency and disposition of ovarian abnor-
Radiologists in ultrasound Consensus Con- malities followed with serial transvaginal ul-
only way to have detected them when ference Statement. Radiology 2010;256: trasonography. Obstet Gynecol 2013;122:
small would have been by screening 943–54. 210–7.
asymptomatic women. Unfortunately, 7. Amor F, Alcázar JL, Vaccaro H, León M, 18. Suh-Burgmann EJ, Alavi M. Detection of
the negative results from randomized Iturra A. GI-RADS reporting system for ultra- early stage ovarian cancer in a large community
controlled trials of ovarian cancer sound evaluation of adnexal masses in clinical cohort. Cancer Med 2019;8:7133–40.
practice: a prospective multicenter study. Ul- 19. Peres LC, Cushing-Haugen KL, Köbel M,
screening indicate that this is not an trasound Obstet Gynecol 2011;38:450–5. et al. Invasive epithelial ovarian cancer survival by
effective strategy. 8. Timmerman D, Testa AC, Bourne T, et al. histotype and disease stage. J Natl Cancer Inst
Simple ultrasound-based rules for the diagnosis 2019;111:60–8.
Conclusion of ovarian cancer. Ultrasound Obstet Gynecol
In this community-based setting, early- 2008;31:681–90.
9. Suh-Burgmann E, Flanagan T, Osinski T, Author and article information
stage high-grade serous cancers rarely Alavi M, Herrinton L. Prospective validation of a From the Division of Gynecologic Oncology, The Perma-
presented as masses of <5 cm or masses standardized ultrasonography-based ovarian nente Medical Group, Oakland, CA (Dr Suh-Burgmann);
without solid areas other than septations cancer risk assessment system. Obstet Gynecol Division of Research, Kaiser Permanente Northern Cali-
on ultrasound examination. Our find- 2018;132:1101–11. fornia, Oakland, CA (Drs Suh-Burgmann and Hung);
ings provide additional support for the 10. Andreotti RF, Timmerman D, Department of Radiology, The Permanente Medical
Strachowski LM, et al. O-RADS US risk stratifi- Group, Oakland, CA (Dr Brasic); and Department of
observation of small masses that do not cation and management system: a consensus Radiology, University of California, San Francisco, San
demonstrate solid areas other than sep- guideline from the ACR ovarian-adnexal report- Francisco, CA (Ms Jha and Dr Goldstein).
tations on ultrasound examination. n ing and data system committee. Radiology Received March 1, 2021; revised April 28, 2021;
2020;294:168–85. accepted April 30, 2021.
References 11. Sharma A, Apostolidou S, Burnell M, et al. This study was funded by the Kaiser Permanente
1. US Cancer Statistics Working Group. US Risk of epithelial ovarian cancer in asymptomatic Northern California Community Benefit Research Pro-
Cancer Statistics Data Visualizations Tool, women with ultrasound-detected ovarian gram and the Permanente Medical Group Physician
based on 2019 submission data (1999-2017): masses: a prospective cohort study within the Researcher Program.
US Department of Health and Human Services, UK collaborative trial of ovarian cancer screening The authors report no conflict of interest.
Centers for Disease Control and Prevention and (UKCTOCS). Ultrasound Obstet Gynecol Corresponding author: Elizabeth Suh-Burgmann, MD.
National Cancer Institute. Centers for Disease 2012;40:338–44. Betty.Suh-Burgmann@kp.org

1.e8 American Journal of Obstetrics & Gynecology MONTH 2021

You might also like