1) What do the authors discuss as the limitations to using polycystic ovarian morphology (PCOM) as part of the diagnostic criteria for polycystic ovarian syndrome (PCOS) in adolescents? What do they state is the purpose of the study?
Answer: PCOM is frequently found in normal adolescence, but can disappear over time. This is attributed to larger ovarian size during adolescence, as anovulation can result in multiple subcentimeter follicles. Follicular counts used in adults may not be applicable in younger patients, and imaging via transabdominal ultrasound can also be limited. The purpose of the study was to determine the frequency of PCOM in healthy adolescent girls with different postmenarchal age ranges, and to determine if a stromal/ovarian area (S/A) ratio is a useful measurement along with ovarian volume and follicle count to label morphology.
2) What were the inclusion and exclusion criteria for the study? What could be some potential limitations with these criteria?
Answer: Subjects were females age 14-18 years with regular menstrual cycles (interval of 21-35 days for at least 3 months) without hirsutism or hyperandrogenemia, and had never had hormonal treatment. Hirsutism was measured by modified Ferriman Gallwey score and acne by the Cremoncini system. Subjects were excluded for invalid pelvic ultrasound, presence of ovarian cysts >3cm on ultrasound, and findings of hyperprolactinemia or elevated androgens on serum screening.
These inclusion and exclusion criteria could be limited by hypothalamic-pituitary-ovarian axis immaturity contributing to cycle intervals falling outside designated inclusion range. Obesity could also impact ability to visualize ovaries well on transabdominal ultrasound. Scoring of acne and hirsutism is subjective, and hair growth can be due to familial patterns or ethnicity and not from hyperandrogenism.
3) Describe the three measurements obtained using pelvic ultrasound, thresholds used for determining PCOM, and how the data was recorded.
Answer: Ovarian volume was calculated with a threshold of ³10cm3. Follicle number per single cross-section (FNPS) and distribution pattern in ovarian maximal plain section was also measured. The threshold of ³12 FNPS was considered PCOM. S/A ratio was measured by outlining the stromal area outer profile with a caliper and also outlining the total area of the ovary with a caliper (see Figure 2). The S/A ratio was calculated by dividing these two parameters, and >0.3 was designated as a cutoff for increased stroma. Transvaginal pelvic ultrasound took the mean calculations of both the right and left ovary. Transabdominal calculations took measurements from only the right ovary.
4) What were the three categories the study authors grouped participants into and how were these defined? What were the statistically significant differences found amongst the 3 groups of participants?
Answer: The study authors stratified participants as either having normal ovarian morphology (NOM), polycystic ovarian morphology with normal stroma (PCOM-NS) or polycystic ovarian morphology with increased stroma (PCOM-IS). They defined NOM as having normal ovarian volume (<10 cm3) with <12 follicles (each <10 mm) OR having a single dominant follicle >10 mm. PCOM was defined as having >12 follicles OR having an ovarian volume >=10 cm3. Normal stroma was defined as S/A <0.3.
There were significantly more subjects with NOM remote from menarche (77% of subjects >5 years from menarche had NOM). PCOM-NS subjects were significantly more common in the first 1-3 years after menarche (57%). The prevalence of PCOM-IS subjects was stable regardless of time from menarche (9-16%). These differences were not correlated to differences in age of menarche, BMI, or waste hip ratio.
LH levels were significantly lower in the NOM group compared to PCOM groups, whether with NS or IS. FSH and E2 levels were similarly low for all three groups, which is expected given that subjects were evaluated in the early follicular phase (5-8 days after LMP).
Although all groups had normal levels of androgens, the PCOM-IS group had relatively higher levels of delta-4-androstenedione (A) compared to the other two groups, and higher levels of total testosterone (tT) than the NOM group.
The PCOM-IS group also had a different follicle distribution pattern. The PCOM-IS group was characterized by a subcortical follicle distribution whereas the other two groups had a diffuse follicle distribution pattern.
5) Based on the findings of the study, the authors extrapolate that it might be possible to use increased ovarian stroma (IS) as an ultrasound finding in adolescents to help distinguish transient changes in polycystic ovarian morphology from permanent alterations. However, the clinical significance of the PCOM-IS morphology can not be concluded from this study. What types of studies would be needed to determine the significance of the PCOM-IS morphology? How would you design such a study/studies?
Because this study only evaluated patients with normal cycles and without evidence of overt hyperandrogenism, it is unclear what the significance of the PCOM-IS findings on ultrasound are in terms of its relation to PCOS. The study authors suggest that because PCOM-IS is associated with relatively higher levels of A and tT, it may more closely resemble a PCOS profile. And because this morphology was found to be stable over the 5 years after menarche, it is unlikely to be transient like the PCOM-NS profile. However given these results alone, it is unclear if this ovarian morphology can potentially change in young adulthood and whether it is predictive for a future diagnosis of PCOS.
One possible study to evaluate the significance of the PCOM-IS morphology would be to prospectively follow subjects who are found to have PCOM-IS within 5 years of menarche and to serially monitor these patients into adulthood. Interval assessments could document if they develop irregular menses and/or hyperandrogenism over time. Interval ultrasounds could also be performed to assess for any changes in ovarian morphology. Such a study could inform us whether PCOM-IS is potentially also a transient morphology from adolescence, but one that resolves later in adulthood, or if it correlates with a diagnosis of PCOS in the future. If subjects with PCOM-IS do go on to meet criteria for PCOS, then this morphology could be used as a predictive tool for PCOS. The drawback to such a study is that it would require years of follow up from study participants and would therefore have a high attrition rate. Additionally, the study would preclude use of hormonal therapies.
Another possible study would be to compare the ultrasound findings of the group of normocyclic, non-hyperandrogenic adolescents to another group with isolated irregular menses but normal androgen levels, and a group with both irregular menses and elevated androgen levels. Ultrasound findings could be compared to see if there are differences in the prevalence of NOM, PCOM-NS, and PCOM-IS in any of the groups. Because irregular menses and elevated androgens in adolescence correlate strongly to PCOS, if this group is also found to have higher rates of PCOM-IS, it may more strongly tie this ovarian morphology to a diagnosis of PCOS. This study would also be more feasible than the previously described hypothetical study given a lack of a need for subjects to follow up long term.