ORIGINAL ARTICLE: REPRODUCTIVE ENDOCRINOLOGY

Impact of right–left differences in

ovarian morphology on the
ultrasound diagnosis of polycystic
ovary syndrome
Brittany Y. Jarrett, Ph.D.,a Heidi Vanden Brink, M.Sc.,a Eric D. Brooks, M.D.,a Kathleen M. Hoeger, M.D.,b
Steven D. Spandorfer, M.D.,c Roger A. Pierson, Ph.D., M.S.,d Donna R. Chizen, M.D.,d
and Marla E. Lujan, Ph.D., M.Sc.a
a
Division of Nutritional Sciences, Cornell University, Ithaca, New York; b Department of Obstetrics and Gynecology, University of
Rochester Medical Center, Rochester, New York; and c Center for Reproductive Medicine, Weill Cornell Medicine, New York,
New York; and d Department of Obstetrics and Gynecology, University of Saskatchewan, Saskatoon, Saskatchewan, Canada

Objective: To assess right–left differences in ultrasonographic markers of ovarian morphology and determine the impact on the
diagnosis of polycystic ovarian morphology (PCOM).
Design: A cross-sectional study of data collected from 2006 to 2018.
Setting: Academic clinical research centers.
Patient(s): Women with polycystic ovary syndrome (PCOS; n ¼ 87) and controls (n ¼ 67).
Intervention(s): None.
Main Outcomes Measure(s): Follicle number per ovary (FNPO), follicle number per cross-section (FNPS), and ovarian volume (OV)
were assessed in both ovaries using transvaginal ultrasonography. PCOM was identified based on recent international consensus
guidelines or proposed diagnostic thresholds.
Result(s): Overall, mean right–left differences were two follicles for FNPO, one follicle for FNPS, and 2 mL for OV. FNPO showed the
strongest correlation between ovaries. Its assessment in a single ovary did not impact the diagnosis of PCOM in women with PCOS.
However, there were differences in the probability of unilateral versus bilateral PCOM based on FNPS and OV in both groups.
Conclusion(s): FNPO is the most reliable unilateral marker of PCOM in light of right–left differences in ovarian morphology. Use of
FNPS or OV to define PCOM is discouraged when only one ovary is visualized. (Fertil SterilÒ 2019;-:-–-. Ó2019 by American
Society for Reproductive Medicine.)
Key Words: Polycystic ovary syndrome, ultrasonography, diagnosis, intraindividual biological variation
Discuss: You can discuss this article with its authors and other readers at https://www.fertstertdialog.com/users/16110-fertility-
and-sterility/posts/49858-28000

A
reliable definition of polycy- syndrome (PCOS) (1–3). Hallmarks of transvaginal ultrasonography (7).
stic ovarian morphology PCOM include follicular excess and The development of higher frequency
(PCOM) is required to identify increased ovarian size (4–6), which ultrasound transducers (i.e.,
clinical variants of polycystic ovary can be observed noninvasively using R8 MHz) has improved the accuracy
of morphological assessments and
Received March 22, 2019; revised May 31, 2019; accepted June 10, 2019. ushered international efforts to
B.Y.J. has nothing to disclose. H.V.B. has nothing to disclose. E.D.B. has nothing to disclose. K.M.H. has
nothing to disclose. S.D.S. has nothing to disclose. R.A.P. is President and CSO of Synergyne Im- re-evaluate the diagnostic markers
aging Technology, Inc. Saskatoon, Saskatchewan Canada. D.R.C. has nothing to disclose. M.E.L. and thresholds for PCOM (3, 7).
has nothing to disclose.
Supported by the Division of Nutritional Sciences at Cornell University, in addition to the President's Follicle number per ovary (FNPO) is
Council of Cornell Women, Saskatchewan Health Research Foundation, Canadian Institutes of now recognized as the most reliable
Health Research, United States Department of Agriculture (grant 2010-34324-20769), and Na-
tional Institutes of Health (grants R56-HD089962 and ULTR00457). B.Y.J. and H.V.B. also received
marker of PCOM (3, 7–14). The
training awards from the National Institutes of Health (grant T32-DK007158) and Canadian In- original definition of follicular excess
stitutes of Health Research (funding reference number 146182). (i.e., FNPO R12) (15, 16) is considered
Reprint requests: Marla E. Lujan, Ph.D., M.Sc., Division of Nutritional Sciences, Cornell University, 216
Savage Hall, Ithaca, New York, 14853 (E-mail: [email protected]). obsolete and experts agree that
thresholds to distinguish normal from
Fertility and Sterility® Vol. -, No. -, - 2019 0015-0282/$36.00
Copyright ©2019 American Society for Reproductive Medicine, Published by Elsevier Inc.
polycystic ovaries (PCO) will require
https://doi.org/10.1016/j.fertnstert.2019.06.016 revision as advances in ultrasound

VOL. - NO. - / - 2019 1

ORIGINAL ARTICLE: REPRODUCTIVE ENDOCRINOLOGY

technology are introduced (3, 7). Modern diagnostic Saskatchewan, Canada) and Institutional Review Boards at
thresholds for FNPO have been proposed by the Androgen Cornell University, University of Rochester, and Weill Cornell
Excess and PCOS Society for its 2014 Task Force Report on Medicine (Ithaca, Rochester, New York, New York). Informed
the Definition and Significance of PCOM (7) and consent was obtained before study procedures were initiated.
International PCOS Network for its 2018 International Some participants have been evaluated in previous publica-
Guideline on the Assessment and Management of PCOS (3). tions (13, 14, 18–21), including one study (13) that
Both groups concluded that the threshold be increased to proposed ultrasonographic thresholds for PCOM.
R25 (7) and R20 follicles per ovary (3), respectively. The
different values reflect the use of different upper limits for
FNPO in a control population—albeit the best approach for Study Participants
establishing a diagnostic threshold is debatable (7). Women of reproductive age (18–38 years) were recruited from
Nevertheless, the acceptance of higher thresholds for FNPO the general population. Women who were pregnant,
is critical to reliably evaluate PCOS with newer imaging lactating, or taking hormonal contraceptives or insulin sensi-
technology. tizers were not eligible to participate in any of the protocols.
Other markers of PCOM (i.e., follicle number per cross- Those with thyroid abnormalities, hyperprolactinemia, his-
section [FNPS] and ovarian volume [OV]) provide inferior tory of oophorectomy, or limited visualization of the ovaries
diagnostic potential for PCOS (8–14). However, FNPS and on ultrasound were excluded from the present study. Women
OV could serve as suitable alternatives to FNPO when the with PCOS (n ¼ 87) were identified by the combined presence
transabdominal route is preferred or image quality is of oligomenorrhea (3) and hyperandrogenism (22). Hyperan-
reduced (3, 7, 14). Notably, FNPS has never been included drogenism was defined as a modified hirsutism score R7 (23)
in international definitions of PCOM (3, 7, 16), but remains or serum total T concentration R61.5 or R127.1 ng/dL (de-
relevant to study given its widespread use as a surrogate pending on the protocol and hormone assay used). Thresholds
marker of follicular excess in clinical practice (7). reflected the 95th percentiles of modified hirsutism score and
Furthermore, the definition of increased ovarian size (i.e., serum total T concentration in a reference cohort. Women
OV R10 mL) has not changed since the Rotterdam with regular cycles and no evidence of hyperandrogenism
consensus (3, 7, 16), suggesting that OV is robust against were included as controls (n ¼ 67).
advances in ultrasound technology.
Despite efforts to refine diagnostic markers and thresh-
olds for PCOM (3, 7, 16), there are limited data on the Ultrasonographic Assessments
appropriateness of technical standards for conducting Participants were evaluated with high-resolution transvagi-
ultrasound assessments. Historically, diagnostic test studies nal ultrasonography. Ultrasonographic data were obtained
have derived thresholds from mean measurements of FNPO, using UltraSonix RP (version 2.3.5; UltraSonix Medical Cor-
FNPS, and OV, with values averaged between the right and poration) and GE Voluson ultrasound machines (E8, S6, or
left ovaries (8–13, 15). Consensus statements have extended S10 Series; GE Healthcare) with 5–9 or 6–12 MHz multifre-
the findings to indicate that unilateral elevation in any quency transducers. Ultrasound examinations were conduct-
ultrasonographic marker is sufficient to meet the definition ed in the early follicular phase of the menstrual cycle
of PCOM (3, 7, 16). The recommendation is presumably (controls) or when no dominant follicles or active corpora lu-
based on early reports that bilateral PCO were the unifying tea (CLs) were present (PCOS). Whole ovaries were scanned
feature of women with PCOS (4, 17). However, application from their inner to outer margins in the longitudinal plane.
of mean thresholds to one side relies on the assumption that Two-dimensional cineloops were archived for offline analysis
there are no measurable, biological differences between the using customized imaging software (Sante DICOM Editor,
right and left ovaries. Few studies have considered right– Santesoft LTD). Images were reviewed by one of three inves-
left differences in ovarian morphology in women of reproduc- tigators. Each investigator demonstrated strong inter-rater
tive age. It stands to reason that lateralization of follicular agreement in FNPO as part of an internal reliability
excess or ovarian size would compromise the diagnostic po- assessment.
tential of PCOM for PCOS, particularly if only one ovary is End points of interest included FNPO, FNPS, and OV.
scanned. Therefore, the present study aimed to assess right– FNPO was assessed throughout each ovary by imposing a pro-
left differences in ultrasonographic markers of ovarian grammable grid onto the viewing window and making
morphology (i.e., FNPO, FNPS, and OV) and determine the focused follicle counts in each grid section (24). FNPS and
impact on the diagnosis of PCOM. OV were obtained in the largest cross-section of each ovary
(14). OV was calculated using the simplified formula for a pro-
late ellipsoid (16). If a cystic structure was detected (e.g., hem-
MATERIALS AND METHODS orrhagic anovulatory follicle, CL, or unspecified cyst), then
Ethical Considerations OV was excluded for both ovaries (n ¼ 5). FNPO, FNPS, and
The present study involved secondary analysis of pooled data OV were tabulated for each ovary. Mean values between
from seven separate protocols. Studies were conducted at four ovaries were calculated and rounded to the nearest whole
clinical research centers in North America from 2006 to 2018. numbers. Sided and mean values were ascribed a morpholog-
Protocols were approved by the Biomedical Research Ethics ical diagnosis based on recent international consensus guide-
Review Board at the University of Saskatchewan (Saskatoon, lines or proposed thresholds for PCOM. PCOM was defined by

2 VOL. - NO. - / - 2019

 Fertility and Sterility®

an FNPO R20 (3), FNPO R25 (7), FNPS R9 (13), or OV similar ages (P ¼ .61), but exhibited higher body mass indices
R10 mL (3, 7, 16). (BMIs), longer menstrual cycles, and higher modified hirsut-
ism scores (all P< .01). The proportion of women with hyper-
Biochemical Assays androgenemia was greater in the PCOS group (38%) versus
the control group (0%) (P< .01). Mean FNPO, FNPS, and OV
Total T was measured with two different liquid chromatog- were elevated in women with PCOS (all P< .01) (Table 1).
raphy tandem-mass spectrometry (LC/MS/MS)-based
methods. The LC/MS/MS was used in combination with
isotope dilution for samples collected at the University of Sas- Intraindividual Variation in FNPO, FNPS, and OV
katchewan (n ¼ 59) (18) and with solid phase extraction for In the entire cohort (n ¼ 154), mean differences between
samples collected at Cornell University, University of Roches- ovaries (i.e., within individuals) were two follicles for FNPO
ter, and Weill Cornell Medicine (n ¼ 95) (21). Intra- assay and (95% confidence interval 0–5; P ¼ .07), one follicle for
interassay coefficients of variation (CV) were <7.5% for both FNPS (95% confidence interval 0–1; P ¼ .01), and 2 mL for
methods. However, the pooled data showed a bimodal distri- OV (95% confidence interval 2–3; P< .01). On average, the
bution, wherein total T concentrations were significantly right ovary contained 5% more follicles and measured 20%
higher in samples treated with isotope-dilution versus solid larger than the left ovary.
phase extraction (Mann-Whitney U test; P< .01). Although The right–left differences in FNPO, FNPS, and OV are pre-
LC/MS/MS is recommended for assessments of total T in sented for each group in Table 2. Mean differences between
women, there can still be considerable variability between ovaries were five follicles for FNPO (P ¼ .02), one follicle
methods and lack of standardization between laboratories for FNPS (P< .01), and 3 mL for OV (P< .01) in women with
(25). To avoid misrepresentation of biochemical data, PCOS, and zero follicles for FNPO (P ¼ .77), zero follicles
participants were categorized as having normoandrogenemia for FNPS (P ¼ .60), and 2 mL for OV (P< .01) in controls. Sig-
or hyperandrogenemia based on assay-specific thresholds nificant correlations were detected between ovaries—across
(described in Study Participants). Only the proportion of markers and within groups. Of the three markers, FNPO
women with biochemical hyperandrogenism is reported showed the strongest correlation between ovaries (PCOS: r
herein. ¼ 0.73, P< .01; controls: r ¼ 0.68, P< .01). The FNPS and
OV markers had moderate correlations between ovaries in
Statistical Analyses women with PCOS (FNPS: r ¼ 0.57, P< .01; OV: r ¼ 0.68,
Statistical analyses were performed with JMP Pro 12 (SAS P< .01), but showed relatively weak relationships in controls
Institute Inc.). The threshold for statistical significance was (FNPS: r ¼ 0.48, P< .01; OV: r ¼ 0.31, P ¼ .01) (Table 2).
set at P< .05. All end points were non-normally distributed.
Demographic and diagnostic characteristics were compared Impact on the Ultrasonographic Diagnosis of PCOS
between groups using Mann-Whitney U or Fisher's exact
tests. Right–left differences in FNPO, FNPS, and OV was as- The proportion of women with unilateral or bilateral PCOM,
sessed using Wilcoxon signed rank tests. Correlations were based on recent international consensus guidelines (3, 7) or
considered with Spearman's rank coefficients. Differences in proposed diagnostic thresholds (13), is shown in Figure 1.
the probability of unilateral versus bilateral PCOM were deter- Most women with PCOS demonstrated bilateral PCOM,
mined using McNemar's test. Sensitivity analyses were also regardless of the ultrasonographic marker or threshold
undertaken in the entire cohort to confirm that findings
were not impacted by methodological or physiological fac-
tors. FNPO, FNPS, and OV were compared across sites, ultra- TABLE 1
sound machines, and maximum transducer frequencies (i.e., 9
vs. 12 MHz) using the Steel-Dwass test. The absence of differ- Demographic and diagnostic characteristics of the study
ences between sites and technologies provided justification participants.
for pooling data across protocols (all P>.10). Mean right– PCOS Control
left differences in FNPO, FNPS, and OV were compared Characteristics (n [ 87) (n [ 67)
between women according to obesity, gravidity, and parity Age (y) 27 (23–30) 27 (23–31)
status using Wilcoxon rank sum tests. The absence of differ- Body mass index (kg/m2) 32.0 (23.7–38.2) 23.6 (21.5–27.2)a
Menstrual cycle length (d) 75 (47–146) 29 (28–32)a
ences between binary categories suggested that study out-
Modified hirsutism score 11 (8–14) 3 (1–5)a
comes were not impacted by the tested physiological factors Proportion with biochemical 33/87 (38%) 0/67 (0%)a
(all P>.05). Post hoc calculations revealed that the present hyperandrogenism
study had 99.9% power to detect the observed intraindividual Mean FNPO 45 (34–60) 25 (16–32)a
Mean FNPS 10 (8–13) 7 (5–9)a
differences in FNPO, FNPS, and OV (a ¼ 0.05). Mean OV (mL) 11 (9–15) 7 (6–9)a
Note: Data are presented as median (interquartile range) or n/total n (percent). Data related
to ultrasonographic features reflect the mean of recorded values from the right and left
RESULTS ovaries rounded to the nearest whole number. FNPO ¼ follicle number per ovary; FNPS ¼
follicle number per cross-section; OV ¼ ovarian volume; PCOS ¼ polycystic ovary syndrome.
Study Participant Characteristics a
Within rows, a significant difference between groups based on the Mann-Whitney U (for
continuous variables) or Fisher's exact test (for categorical variables), P< .05.
Demographic and diagnostic characteristics are reported in
Jarrett. Right–left differences in PCOM. Fertil Steril 2019.
Table 1. Compared with controls, women with PCOS were of

VOL. - NO. - / - 2019 3

ORIGINAL ARTICLE: REPRODUCTIVE ENDOCRINOLOGY

not FNPO. The present study is timely given increasing global

TABLE 2
attention to the definition of PCOM and its implications for
Right–left differences in ultrasonographic features in controls and
evidence-based guidelines to diagnose and manage PCOS
women with PCOS. (3, 7). Data reported herein are unique in that we used
Right Left MD ± SE higher frequency ultrasound transducers (7) and reliable
Characteristics ovary ovary (95% CI) Correlation measurement techniques (24) in well-defined cohorts.
PCOS (n ¼ 87)
We propose that FNPO is the most reliable intraindividual
FNPO 45 (33–69) 43 (32–54) 5  2 (1, 8)a 0.73a marker of PCOM. The FNPO showed the strongest correlation
FNPS 11 (8–13) 10 (7–12) 1  0 (0, 2)a 0.57a between ovaries and there were no differences in the proba-
OV 13 (9–17) 10 (7–14) 3  1 (2, 4)a 0.59a bility of unilateral versus bilateral PCOM in women with
Control (n ¼ 67)
FNPO 23 (16–32) 24 (15–33) 0  1 (–2, 2) 0.68a PCOS. Our findings add to evidence that FNPO offers the
FNPS 7 (4–10) 7 (5–9) 0  0 (–1, 1) 0.48a most diagnostic potential for PCOS (3, 7–14). Interestingly,
OV 8 (6–11) 6 (4–9) 2  1 (1, 3)a 0.31a 1 in 10 controls had elevated follicle counts in a single
Note: Data are presented as median (interquartile range) unless otherwise specified and ovary by both the International PCOS Network and
reflect recorded values from individual ovaries rounded to the nearest whole number. The
mean difference (MD) reflects the right minus the left ovary. Therefore, a positive number Androgen Excess and PCOS Society criteria (3, 7). Proximity
reflects a larger value in the right ovary. CI ¼ confidence interval; FNPO ¼ follicle number
per ovary; FNPS ¼ follicle number per cross-section; OV ¼ ovarian volume; SE ¼ standard of follicle counts to the diagnostic threshold (i.e., within 0–2
error.
a
follicles) appeared to contribute to differences in the
A significant mean difference (based on Wilcoxon signed rank tests) or correlation (based
on Spearman's rank coefficient) between ovaries, P< .05. probability of unilateral versus bilateral PCOM in controls.
Jarrett. Right–left differences in PCOM. Fertil Steril 2019. Therefore, we suspect that right–left differences in FNPO are
most relevant to the diagnosis of PCOM in women with
mild follicular excess (i.e., 15–30 follicles).
applied (Fig. 1). The FNPO represented the most consistent FNPS and OV seem to be less reliable intraindividual
intraindividual marker of PCOM in PCOS; 94% of women markers of PCOM. Both end points had moderate correlations
had R20 follicles in both ovaries (Fig. 1A) and 89% had between ovaries in women with PCOS, but weak correlations
R25 follicles in both ovaries (Fig. 1B). Less than 4% of in controls. Discordant diagnoses of PCOM occurred when
women with PCOS had elevated FNPO in a single ovary by FNPS and OV were only considered in a single ovary. Previous
either diagnostic threshold (Fig. 1A and B). No differences studies have also reported weaker correlations between
were observed in the probability of unilateral versus ovaries for OV versus FNPO (26) and suggested that size dif-
bilateral PCOM based on FNPO R20 (k ¼ 0.88; P ¼ .32) or ferences make PCOM more likely on the right side (27). As a
R25 (k ¼ 0.86; P ¼ .16). The FNPS and OV were less result Atiomo and colleagues (28) proposed that the right
consistent intraindividual markers of PCOM in PCOS. ovary could offer differential diagnostic potential for PCOS.
Although two-thirds of women with PCOS demonstrated Post-hoc analyses of our data revealed that right OV (area un-
bilateral PCOM with each marker, 16% had elevated FNPS der the receiver operating characteristic curve ¼ 0.78; P< .05)
in a single ovary (Fig. 1C) and 12% had elevated OV in a single and left OV (area under the receiver operating characteristic
ovary (Fig. 1D). Differences were observed in the probability curve ¼ 0.78; P< .05) offered similar diagnostic potential
of unilateral versus bilateral PCOM based on FNPS R9 (k ¼ for PCOS. However, a higher threshold for OV was required
0.59; P< .01) and OV R10 mL (k ¼ 0.70; P< .01). in the right ovary (threshold: 10 mL; sensitivity: 74%; speci-
Most women in the control group also demonstrated ficity: 71%) versus the left ovary (threshold: 9 mL; sensitivity:
bilateral PCOM based on FNPO; 67% of women had R20 fol- 67%; specificity: 76%) to distinguish women with PCOS from
licles in both ovaries (Fig. 1A) and 52% had R25 follicles in controls. We appreciate that generating side-specific diag-
both ovaries (Fig. 1B). FNPO continued to represent the most nostic thresholds for PCOM would lead to confusion in clin-
consistent intraindividual marker in controls; only 9%–11% ical practice. Therefore, because FNPS and OV may be most
of women had variable morphology between ovaries prone to right–left differences, we recommend that their use
(Fig. 1A and B). Fewer controls demonstrated bilateral be avoided if only one ovary can be assessed.
PCOM based on FNPS R9 (30%) or OV R10 mL (24%) and We noted that the right ovary contained more follicles
19% had elevated FNPS (Fig. 1C) or elevated OV in a single than the left ovary in women with PCOS, consistent with
ovary (Fig. 1D). Such variation contributed to differences in the greater volume of the right ovary. Our findings are similar
the probability of unilateral versus bilateral PCOM based on to those of some (26, 27, 29, 30), but not all (31), previous
FNPO R20 (k ¼ 0.79; P ¼ .01), FNPO R25 (k ¼ 0.79; studies of right–left differences in FNPO in women with infer-
P< .01), FNPS R9 (k ¼ 0.61; P< .01), and OV R10 mL (k ¼ tility or PCOS, where the investigators (26, 27, 29, 30) have
0.59; P< .01) (Fig. 1). reported 0.1–2.0 more follicles in the right ovary. Our
higher result of five follicles could reflect our unique
approach to offline image analysis (24) compared with other
DISCUSSION investigators' real-time (27, 29–31) or semiautomated
We detected right–left differences in ultrasonographic counts (26). There are some challenges associated with the
markers of ovarian morphology, which had a significant assessment of FNPO, including distinguishing an anechoic
impact on the identification of PCOM, depending on the ovary structure as one or two clustered follicles and recounting
and marker considered. Discordant diagnoses of PCOM in one follicles that have already been flagged. It is especially
versus both ovaries occurred with the use of FNPS and OV, but difficult to obtain reliable counts amidst the follicular

4 VOL. - NO. - / - 2019

 Fertility and Sterility®

FIGURE 1

Proportion of women with unilateral or bilateral polycystic ovarian morphology. Women with polycystic ovary syndrome (PCOS) (n ¼ 87) and
controls (n ¼ 67) were categorized based on the presence of polycystic ovarian morphology in neither (light gray), one (dark gray), or both
ovaries (black). Polycystic ovarian morphology was defined according to recent international consensus guidelines or proposed diagnostic
thresholds including: (A) follicle number per ovary (FNPO) R20 (3), (B) FNPO R25 (7), (C) follicle number per cross-section (FNPS) R9 (13),
and (D) ovarian volume (OV) R10 mL (3, 7, 16).
Jarrett. Right–left differences in PCOM. Fertil Steril 2019.

density in PCO (32, 33). By evaluating each ovary in small unspecified infertility have suggested that right–left
sections, we improve reproducibility and minimize the differences in FNPO reflect an increased number of small-
likelihood of undercounting or overcounting follicles (24). sized (27) or medium-sized (26) follicles in the right ovary.
Although our approach is not feasible in clinical practice It could be speculated that aspects of follicular recruitment
due to the time required for analysis, it offers a level of vary between the two ovaries. This observation has not
accuracy that is beneficial in research settings (24). Use of been made in healthy women (40), but longitudinal studies
the grid system may also explain why we did not detect could reveal the mechanisms behind right–left differences
right-sided differences in FNPO in controls (i.e., mean differ- in FNPO in PCOS.
ence, 0 follicles) when other investigators (29, 34) did (i.e., We also noted that the right ovary was larger than the
0.4–1.0 follicles). left ovary in both groups. Our findings are consistent with
It is important to consider the clinical relevance of intra- recent studies of right–left differences in OV in women with
individual differences in follicle number. Deb et al. (35, 36) regular cycles (34), infertility (26, 27, 41), and PCOS (29).
have reported intraobserver and intermethod variability in Average right–left differences have ranged from 0.5–
counts as wide as 11 follicles, suggesting that mean 1.6 mL (26, 27, 29, 34, 41), with one study (26) reporting
differences <11 could be within error. Small differences an upper limit of agreement near 5.0 mL. Evidence that
would also be difficult to capture in a single cross-section the right ovary is larger than the left ovary across the
of the ovary (i.e., FNPS). Alternatively, evidence of more lifespan, from childhood (42) through the fourth decade
frequent natural and stimulated ovulations in the right ovary (27, 29) and postmortem (43), increases confidence in our
(37–39) supports the notion that ovarian reserve (and thus, findings. Differences in blood supply between the two
folliculogenesis) is lateralized. Studies in cohorts with ovaries could offer a physiological basis for one being

VOL. - NO. - / - 2019 5

ORIGINAL ARTICLE: REPRODUCTIVE ENDOCRINOLOGY

larger—albeit studies (26, 34, 41, 44, 45) using Doppler nosed in a single ovary using FNPO, but not FNPS or OV. We
ultrasound technology have yielded conflicting data on recommend that practitioners and researchers acknowledge
right–left differences in vascular indices, independent of the potential for right–left differences in ultrasonographic
gravidity or parity. markers of ovarian morphology and use FNPO to define
An unexpected finding was that 52%–67% of controls PCOM when only one ovary can be visualized.
had bilateral PCOM based on modern diagnostic thresholds
for FNPO. Our observation is reminiscent of concerns that
emerged following publication of the Rotterdam criteria in
2003, when a heightened prevalence of PCOM was described REFERENCES
in healthy women of reproductive age (46). We acknowledge 1. Rotterdam ESHRE/ASRM-sponsored PCOS Consensus Workshop Group.
that use of the grid system (24) could lead to increased detec- Revised 2003 consensus on diagnostic criteria and long-term health
tion of smaller follicles (e.g., 2–3 mm)—to an extent that may risks related to polycystic ovary syndrome (PCOS). Hum Reprod 2004;
limit the external validity of our results. When we excluded 2- 19:41–7.
to 3-mm follicles from control estimates of FNPO, the median 2. Azziz R, Carmina E, Dewailly D, Diamanti-Kandarakis E, Escobar-Morreale H,
Futterweit W, et al. Position Statement: criteria for defining polycystic ovary
value decreased from 25–14 follicles and fewer controls had
syndrome as a predominantly hyperandrogenic syndrome: an Androgen
PCOM based on FNPO R20 (16%) or R25 (3%) (data not Excess Society guideline. J Clin Endocrinol Metab 2006;91:4237–45.
shown). We agree that standardized methods and training 3. Teede H, Misso M, Costello M, Dokras A, Laven J, Moran L, et al. Recom-
are needed to obviate discrepancies in follicle counts among mendations from the international evidence-based guideline for the assess-
studies and in clinical practice (3, 32, 33). Such efforts ment and management of polycystic ovary syndrome. Fertil Steril 2018;110:
could be deprioritized if more reliable assays are developed 364–79.
and surrogate markers of PCOM (e.g., antim€ ullerian 4. Hughesdon P. Morphology and morphogenesis of the Stein-Leventhal ovary
and of so-called ‘‘hyperthecosis’’. Obstet Gynecol Surv 1982;37:59–77.
hormone) are deemed effective replacements for FNPO
5. Maciel G, Baracat E, Benda J, Markham S, Hensinger K, Chang R, et al.
(3, 7). It should be stated that PCOS cannot be diagnosed Stockpiling of transitional and classic primary follicles in ovaries of women
with a single feature (1–3) and it is unclear whether PCOM with polycystic ovary syndrome. J Clin Endocrinol Metab Metab 2004;89:
confers health risks independent of anovulation or 5321–7.
androgen excess. Nevertheless, PCOM increases the 6. Webber L, Stubbs S, Stark J, Trew G, Margara R, Hardy K, et al. Formation
likelihood of ovarian hyperstimulation syndrome (OHSS) and early development of follicles in the polycystic ovary. Lancet 2003;
during gonadotropin therapy and has implications for 362:1017–21.
7. Dewailly D, Lujan M, Carmina E, Cedars M, Laven J, Norman R, et al. Defi-
success of treatment (47). Studies that characterize the
nition and significance of polycystic ovarian morphology: A task force report
physiological range of FNPO and OV would certainly help from the androgen excess and polycystic ovary syndrome society. Hum Re-
to clarify the significance of PCOM in women with and prod Update 2014;20:334–52.
without PCOS. 8. Jonard S, Robert Y, Dewailly D. Revisiting the ovarian volume as a diagnostic
The present study had two limitations. First, in an effort to criterion for polycystic ovaries. Hum Reprod 2005;20:2893–8.
increase our sample size, we pooled ultrasonographic data ob- 9. Allemand M, Tummon I, Phy J, Foong S, Dumesic D, Session D. Diagnosis of
polycystic ovaries by three-dimensional transvaginal ultrasound. Fertil Steril
tained at four sites with two brands of ultrasound machines.
2006;85:214–9.
We used the same acquisition protocols across sites and relied 10. Chen Y, Li L, Chen X, Zhang Q, Wang W, Li Y, et al. Ovarian volume and fol-
on newer technology (i.e., as defined by a maximum trans- licle number in the diagnosis of polycystic ovary syndrome in Chinese
ducer frequency R8 MHz). However, we appreciate that women. Ultrasound Obstet Gynecol 2008;32:700–3.
combining multiple datasets can introduce error and 11. Ko€ş€ €ş€
uş N, Ko uş A, Turhan N, Kamalak Z. Do threshold values of ovarian vol-
confound the true effect(s) of interest. We did not detect an ume and follicle number for diagnosing polycystic ovarian syndrome in Turk-
impact of site or technology on FNPO, FNPS, or OV, but ish women differ from western countries? Eur J Obstet Gynecol Reprod Biol
2011;154:177–81.
recommend that our findings be interpreted with caution.
12. Dewailly D, Gronier H, Poncelet E, Robin G, Leroy M, Pigny P, et al. Diagnosis
Second, the present study was not designed to identify the of polycystic ovary syndrome (PCOS): revisiting the threshold values of folli-
factors responsible for right–left differences in ovarian cle count on ultrasound and of the serum AMH level for the definition of
morphology. Although we accounted for the potential influ- polycystic ovaries. Hum Reprod 2011;26:3123–9.
ence of obesity, gravidity, and parity, we acknowledge that 13. Lujan M, Jarrett B, Brooks E, Reines J, Peppin A, Muhn N, et al. Updated ul-
the ovaries are more difficult to visualize in women with se- trasound criteria for polycystic ovary syndrome: reliable thresholds for
vere obesity and that our inclusion criteria may have skewed elevated follicle population and ovarian volume. Hum Reprod 2013;28:
1361–8.
our assessments to leaner cohorts. Likewise, our use of the Na-
14. Christ J, Willis A, Brooks E, Vanden Brink H, Jarrett B, Pierson R, et al. Follicle
tional Institutes of Health criteria precluded us from evalu- number, not assessments of the ovarian stroma, represents the best
ating any role of phenotype. Use of the Rotterdam criteria ultrasonographic marker of polycystic ovary syndrome. Fertil Steril 2014;
in future studies would help to elucidate the effect(s) of anov- 101:280–7.
ulation and hyperandrogenism on right–left differences in 15. Jonard S, Robert Y, Cortet-Rudelli C, Pigny P, Decanter C, Dewailly D. Ultra-
FNPO, FNPS, and OV. sound examination of polycystic ovaries: is it worth counting the follicles?
Hum Reprod 2003;18:598–603.
In conclusion, significant intraindividual differences
16. Balen A, Laven J, Tan S, Dewailly D. Ultrasound assessment of the polycystic
were observed in ultrasonographic markers of ovarian ovary: international consensus definitions. Hum Reprod Update 2003;9:
morphology among controls and women with PCOS. FNPO 505–14.
showed the smallest differences between ovaries. Our data 17. Stein I, Leventhal M. Amenorrhea associated with bilateral polycystic ovaries.
may be interpreted to mean that PCOM can be reliably diag- Am J Obstet Gynecol 1935;29:181–91.

6 VOL. - NO. - / - 2019

 Fertility and Sterility®

18. Lujan M, Bloski T, Chizen D, Lehotay D, Pierson R. Digit ratios do not serve as 33. Lujan M, Chizen D, Peppin A, Dhir A, Pierson R. Assessment of
anatomical evidence of prenatal androgen exposure in clinical phenotypes ultrasonographic features of polycystic ovaries is associated with modest
of polycystic ovary syndrome. Hum Reprod 2010;25:204–11. levels of inter-observer agreement. J Ovarian Res 2009;2:6.
19. Christ J, Vanden Brink H, Brooks E, Pierson R, Chizen D, Lujan M. Ultra- 34. Jokubkiene L, Sladkevicius P, Valentin L. Number of antral follicles, ovarian
sound features of polycystic ovaries relate to degree of reproductive and volume, and vascular indices in asymptomatic women 20 to 39 years old
metabolic disturbance in polycystic ovary syndrome. Fertil Steril 2015;103: as assessed by 3-dimensional sonography. J Ultrasound Med 2012;31:
787–94. 1635–49.
20. Clark N, Podolski A, Brooks E, Chizen D, Pierson R, Lehotay D, et al. Preva- 35. Deb S, Campbell B, Clewes J, Raine-Fenning N. Quantitative analysis of
lence of polycystic ovary syndrome phenotypes using updated criteria for antral follicle number and size: a comparison of two-dimensional and auto-
polycystic ovarian morphology: an assessment of over 100 consecutive mated three-dimensional ultrasound techniques. Ultrasound Obstet Gyne-
women self-reporting features of polycystic ovary syndrome. Reprod Sci col 2010;35:354–60.
2014;21:1034–43. 36. Deb S, Jayaprakasan K, Campbell B, Clewes J, Johnson I, Raine-Fenning N.
21. Vanden Brink H, Willis A, Jarrett B, Lin A, Soler S, Best S, et al. Sonographic Intraobserver and interobserver reliability of automated antral follicle counts
markers of ovarian morphology, but not hirsutism indices, predict serum to- made using three-dimensional ultrasound and SonoAVC. Ultrasound Obstet
tal testosterone in women with regular menstrual cycles. Fertil Steril 2016; Gynecol 2009;33:477–83.
105:1322–9. 37. Fukuda M, Fukuda K, Yding Andersen C, Byskov AG. Does corpus
22. Zawadzki J, Dunaif A. Diagnostic criteria for polycystic ovary syn- luteum locally affect follicular growth negatively? Hum Reprod 1997;12:
drome: towards a rational approach. In: Dunaif A, Givens J, 1024–7.
Haseltine F, editors. Polycystic ovary syndrome. Boston: Blackwell 38. Fukuda M, Fukuda K, Yding Andersen C, Byskov A. Right-sided ovulation fa-
Scientific; 1992:377–84. vours pregnancy more than left-sided ovulation. Hum Reprod 2000;15:
23. Yildiz B, Bolour S, Woods K, Moore A, Azziz R. Visually scoring hirsutism. 1921–6.
Hum Reprod Update 2009;16:51–64. 39. Lan K, Huang F, Lin Y, Kung F, Lan T, Chang S. Significantly superior
24. Lujan M, Brooks E, Kepley A, Chizen D, Pierson R, Peppin A. Grid analysis im- response in the right ovary compared with the left ovary after stimulation
proves reliability in follicle counts made by ultrasonography in women with with follicle-stimulating hormone in a pituitary down-regulation regimen.
polycystic ovary syndrome. Ultrasound Med Biol 2010;36:712–8. Fertil Steril 2010;93:2269–73.
25. Rosner W, Auchus R, Azziz R, Sluss P, Raff H. Position statement: utility, 40. Baerwald A, Adams G, Pierson R. A new model for ovarian follicular devel-
limitations, and pitfalls in measuring testosterone: an Endocrine Society Po- opment during the human menstrual cycle. Fertil Steril 2003;80:116–22.
sition Statement. J Clin Endocrinol Metab 2007;92:405–13. 41. Jarvela I, Mason H, Sladkevicius P, Kelly S, Ojha K, Campbell S, et al. Char-
26. Deb S, Kannamannadiar J, Campbell B, Clewes J, Raine-Fenning N. The in- acterization of normal and polycystic ovaries using three-
terovarian variation in three-dimensional ultrasound markers of ovarian dimensional power Doppler ultrasonography. J Assist Reprod Genet 2002;
reserve in women undergoing baseline investigation for subfertility. Fertil 19:582–90.
Steril 2011;95:667–72. 42. Griffin I, Cole T, Duncan K, Hollman A, Donaldson M. Pelvic ultrasound
27. Korsholm A-S, Hvidman H, Bentzen J, Andersen A, Petersen K. Left–right dif- measurements in normal girls. Acta Pædiatrica 1995;84:536–43.
ferences in ovarian volume and antral follicle count in 1423 women of repro- 43. Perven H, Nurunnabi A, Ara S, Jahan M. Cadaver study of the volume of
ductive age. Gynecol Endocrinol. 2016. Available at: https://www. the ovary in Bangladeshi women. Bangladesh Med Res Counc Bull 2014;
tandfonline.com/doi/full/10.1080/09513590.2016.1259406. Accessed 40:15–8.
December 12, 2016. 44. Pan H, Wu M, Cheng Y, Li C, Chang F. Quantification of Doppler signal in
28. Atiomo W, Pearson S, Shaw S, Prentice A, Dubbins P. Ultrasound criteria in polycystic ovary syndrome using three-dimensional power Doppler
the diagnosis of polycystic ovary syndrome (PCOS). Ultrasound Med Biol ultrasonography: a possible new marker for diagnosis. Hum Reprod 2002;
2000;26:977–80. 17:201–6.
29. Ko€ninger A, Koch L, Edimiris P, Nießen S, Kasimir-Bauer S, Kimmig R, et al. 45. Wu M, Tang H, Hsu C, Wang S, Huang K. The role of three-dimensional ultrasono-
Intraindividual right–left comparison of sonographic features in polycystic graphic images in ovarian measurement. Fertil Steril 1998;69:1152–5.
ovary syndrome (PCOS) diagnosis. Eur J Obstet Gynecol 2014;181:124–9. 46. Johnstone E, Rosen M, Neril R, Trevithick D, Sternfeld B, Murphy R, et al. The
30. Alserri A, Kuriya A, Holzer H, Tulandi T. Lateralization of ovarian follicles. Gy- polycystic ovary post-Rotterdam: a common age-dependent finding in
necol Obstet Invest 2014;77:117–20. ovulatory women without metabolic significance. J Clin Endocrinol Metab
31. Pache T, Wladimiroff J, Hop W, Fauser B. How to discriminate between 2010;95:4965–72.
normal and polycystic ovaries: transvaginal US study. Radiology 1992;183: 47. Hendriks D, Mol B, Bancsi L, te Velde E, Broekmans F. Antral follicle
421–3. count in the prediction of poor ovarian response and pregnancy
32. Lujan M, Chizen D, Peppin A, Kriegler S, Leswick D, Bloski T, et al. Improving after in vitro fertilization: a meta-analysis and comparison
inter-observer variability in the evaluation of ultrasonographic features of with basal follicle-stimulating hormone level. Fertil Steril 2005;83:
polycystic ovaries. Reprod Biol Endocrinol 2008;6:30. 291–301.

VOL. - NO. - / - 2019 7