Renaming PCOS—A Two-State Solution

Abstract

Context:

It has become evident over the past 30 years that polycystic ovary syndrome (PCOS) is more than a reproductive disorder. It has metabolic sequelae that can affect women across the lifespan. Diagnostic criteria based on the endocrine features of the syndrome, hyperandrogenism and chronic anovulation, such as the National Institutes of Health (NIH) criteria, identify women at high metabolic risk. The additional phenotypes defined by the Rotterdam diagnostic criteria identify women with primarily reproductive rather than metabolic dysfunction.

Objective:

The aim is to discuss the rationale for a separate name for the syndrome that is associated with high metabolic risk while maintaining the current name for the phenotypes with primarily reproductive morbidity.

Intervention:

The NIH Office for Disease Prevention-Sponsored Evidence-Based Methodology Workshop on Polycystic Ovary Syndrome recommended that a new name is needed for PCOS.

Positions:

The authors propose that PCOS be retained for the reproductive phenotypes and that a new name be created for the phenotypes at high metabolic risk.

Conclusions:

There should be two names for the PCOS phenotypes: those with primarily reproductive consequences should continue to be called PCOS, and those with important metabolic consequences should have a new name.

The National Institutes of Health (NIH) Office for Disease Prevention (ODP) sponsored the first Evidence-Based Methodology Workshop on Polycystic Ovary Syndrome on December 3–5, 2012. The conference followed a standard format for meetings organized by ODP to address controversial medical issues. The state-of-the-science in the field was presented to an impartial panel of experts in endocrinology, metabolism, cardiology, epidemiology, obstetrics, gynecology, and primary care who were not investigators in polycystic ovary syndrome (PCOS). The panel was asked to clarify: 1) the benefits and drawbacks of different diagnostic criteria; 2) the causes, predictors, and long-term consequences of PCOS; and 3) optimal prevention and treatment strategies. All previous meetings on PCOS diagnostic criteria (1–4), despite being labeled consensus conferences, have not followed any formal consensus process but have resulted in recommendations based on expert opinion, the lowest level of scientific evidence (5, 6).

The panel made a number of detailed and cogent recommendations that can be viewed in their report at http://prevention.nih.gov/workshops/2012/pcos/resources.aspx. The major recommendation on which we would like to comment as the lead organizers of the 1990 National Institute of Child Health and Human Development (NICHD) and the 2003 European Society of Human Reproduction and Embryology (ESHRE)-American Society of Reproductive Medicine (ASRM) Rotterdam conferences on PCOS is the proposal to change the name of the syndrome. Specifically, we want to ensure that this recommendation does not lead to Balkanization of the field, which will clearly undermine the broad interdisciplinary efforts required for meaningful scientific advances in our understanding of PCOS.

We propose a nosological “two-state solution” to the conflict. The endocrine syndrome of hyperandrogenism and chronic anovulation, eg, the NIH phenotype, should have a new name that acknowledges both its reproductive features and its long-term metabolic risks. The phenotypes diagnosed by ovarian morphology, eg, the remaining Rotterdam phenotypes, should continue to be known as PCOS. The concept of identifying the high-risk NIH phenotype with a second name was supported at the most recent ESHRE-ASRM meeting of PCOS experts in Amsterdam in 2010 (7, 8). Furthermore, this proposal would be consistent with the NIH expert panel's recommendation that the Rotterdam diagnostic criteria should continue to be used.

The nosological difficulties presented by PCOS are not new. They were most poetically characterized by Netter et al (9) who wrote in 1958 that “the syndrome of Stein (as PCOS was previously known) is a fugitive syndrome, with limits less well defined than those of the Sahara or the Sudan.” However, the first attempt to develop standardized diagnostic criteria for the syndrome did not come until 1990 at a conference sponsored by NICHD. The resulting NIH criteria were those features of the syndrome receiving the most votes in an informal poll of the conference attendees who were experts in the field (1). The adoption of these criteria resulted in a standardized classification of patients in research greatly facilitating the characterization of the reproductive and metabolic features of PCOS over the ensuing two decades (10).

Nevertheless, the NIH criteria were immediately controversial because polycystic ovarian morphology (PCO) was not included as a criterion since evidence suggested even then that the morphologic changes were not specific (11). However, outside the United States, ovarian ultrasound imaging was a standard procedure for the diagnostic evaluation of potentially affected women. Moreover, with the increasing use of assisted reproductive technologies in which ovarian imaging is an essential component of care, it became evident that women with PCO were at increased risk for ovarian hyperstimulation syndrome (12).

These concerns led to the 2003 ESHRE-ASRM-sponsored Rotterdam conference (2, 3) that re-evaluated the diagnostic criteria and recommended the addition of PCO as a diagnostic criterion to the NIH criteria. However, the Rotterdam criteria required only two of the three diagnostic criteria (Table 1) for the diagnosis of PCOS so that they encompassed the NIH criteria. This change resulted in two additional phenotypes, ie, women with PCO and oligo/amenorrhea, and women with hyperandrogenism and PCO (Table 2). In 2006, an expert panel of the Androgen Excess Society (4) recommended a modification of the Rotterdam criteria to require hyperandrogenism as an essential diagnostic criterion (Table 1), thereby reducing the number of PCOS phenotypes from three to two (Table 2). However, the AES criteria have not been widely adopted.

Table 1.

Diagnostic Criteria for PCOS

Criteria
NIH--NICHD (1)	Both hyperandrogenism and chronic anovulation
Rotterdam (2, 3)	Two of the following: hyperandrogenism, chronic anovulation, and polycystic ovaries
Androgen Excess Society (4)	Hyperandrogenism plus ovarian dysfunction indicated by oligo/amenorrhea and/or polycystic ovaries

All criteria require exclusion of other disorders: hyperprolactinemia, nonclassic congenital adrenal 21-hydroxylase deficiency, thyroid dysfunction, androgen-secreting neoplasms, and Cushing's syndrome.

Table 2.

PCOS Phenotypes According to Diagnostic Criteria Applied

	HA and Anov	HA and PCO	Anov and PCO	HA	PCO	Anov
NICHD	+	−	−	−	−	−
Rotterdam	+	+	+	−	−	−
AES	+	+	−	−	−	−

Abbreviations: HA, hyperandrogenism; Anov, chronic anovulation; PCO, polycystic ovarian morphology.

The broadening of the diagnostic criteria has led to a large body of research comparing the PCOS phenotypes (for reviews, see Refs. 10, 13, and 14). However, an unintended consequence of the broadening of the diagnostic criteria has been the inclusion in studies of multiple PCOS phenotypes without stratification. This failure to investigate precisely defined PCOS phenotypes has resulted in confusion in the literature because the metabolic features of the syndrome, in particular, vary by phenotype (10, 14, 15). Improvements in ultrasound technology have led to revisions in criteria for PCO (16). Further studies have found that polycystic ovarian changes are age-related and are present in a large number of otherwise reproductively and metabolically normal women (17). The limitations of total and free T assays for the diagnosis of hyperandrogenism in women are now well-recognized (18). Despite all of these limitations, replicated genetic loci have been associated with both the NIH and Rotterdam PCOS phenotypes (19–22).

There is now substantial evidence from cross-sectional studies that the metabolic features of the syndrome differ by phenotype—the NIH phenotype is at high risk for insulin resistance and its associated features, metabolic syndrome and type 2 diabetes (reviewed in Refs. 10, 13, and 14). These metabolic abnormalities are less severe or are absent in the other phenotypes (14, 15), particularly those women with regular menses (15, 23). Although prospective studies are essential to determine the long-term health risks of the various phenotypes, there was agreement among experts at the 2010 Amsterdam conference that the NIH criteria identified the metabolically high-risk group (7, 8). However, it is important to ensure that all research on PCOS classifies women by phenotype.

A major barrier to the effective comprehensive healthcare of women with PCOS is the fact that this syndrome is little recognized outside of obstetrics and gynecology. Indeed, a population-based study found that almost 70% of women fulfilling either NIH or Rotterdam criteria had previously undiagnosed PCOS (24). Accordingly, affected women at high risk for prediabetes, type 2 diabetes, and other cardiometabolic risk factors are not identified by pediatricians, internists, and family practitioners—those specialties primarily responsible for long-term risk reduction interventions. In addition, misperception of risk can reduce patient compliance with lifestyle modification. One can speculate that the focus of the current name on ovarian morphology is a deterrent to both healthcare providers and patients in recognizing the metabolic risks of the syndrome.

A central issue in increasing the awareness of PCOS beyond the subspecialty community is the availability of and familiarity with the requisite diagnostic procedures. Diagnostic criteria that require ovarian ultrasonography constrain the diagnosis of the syndrome by healthcare providers who do not perform their own ovarian ultrasounds because general radiologists are often unfamiliar with the criteria for PCO and, thus, rarely diagnose this change. Moreover, cross-sectional (17) as well as prospective studies (25) indicate that ovarian morphology itself does not identify those women at metabolic risk. In contrast, the preponderance of the evidence indicates that the endocrine features of the syndrome, hyperandrogenism and oligoanovulation, are sufficient for identifying women at high risk for metabolic abnormalities (8, 10, 13–15, 23).

Indeed, a simple menstrual history may be extremely valuable for detection of underlying medical disorders (26). Approximately 80% of such women present with normal estradiol levels (also referred to as World Health Organization [WHO] group 2 anovulation) and PCOS. In contrast, women with hypoestrogenic oligo/amenorrhea (due to either ovarian insufficiency [WHO group 3] or hypothalamic-pituitary dysfunction [WHO group 1]) present with other health risks, such as loss of bone mass. Therefore, disseminating the concept that menses are a vital sign, as has been recommended by American Academy of Pediatrics Committee on Adolescence and the American College of Obstetricians and Gynecologists Committee on Adolescent Health Care (26), will lead to the identification of women at risk for adverse health outcomes for which we have effective interventions. T assays have their limitations, but the Centers for Disease Control is undertaking an initiative to standardize these assays (27), similar to what was done with cholesterol measurements, so this limitation should be resolved in the near future.

It is clearly feasible for generalists to make the diagnosis of the endocrine features of PCOS, hyperandrogenism, and oligoanovulation with simple education once they are motivated by an appreciation of the health implications of the disorder. Diagnostic criteria, such as the NIH criteria, that do not require ovarian morphology are the most practical for application in the primary care setting, coupled with a new name that does not include the ovarian morphological features of the syndrome. New names for disorders have a checkered past. The proposal to change the name of PCOS to chronic hyperandrogenic anovulation never gained traction. In contrast, identifying the constellation of cardiometabolic risk factors as “metabolic syndrome” has been embraced (28). Therefore, perhaps a related name, the “metabolic reproductive syndrome” (MRS) (as suggested by Sheila E. Laredo, MD, PhD, University of Toronto), would be successful.

Obstetricians and gynecologists, particularly those specializing in reproductive endocrinology and infertility, are aware that women with PCO, independent of hormonal changes, are at high risk for ovarian hyperstimulation, a potentially fatal complication of exogenous gonadotropin therapy for infertility (12). Accordingly, the continued focus on ovarian morphology is appropriate among specialists treating infertility in affected women. The role of insulin resistance in the pathogenesis of PCOS appears to be well-recognized among obstetricians and gynecologists as indicated by the widespread use of insulin-sensitizing drugs, in particular metformin, as a component of assisted reproductive therapies in affected women (29, 30).

Collaborative efforts with healthcare professionals across a broad range of specialties providing care for women as well as with patients in the model pioneered in Australia (www.managingpcos.org.au) will be needed for the complex reproductive and metabolic sequelae of PCOS to be managed effectively across the lifespan. It is essential to recognize the perspectives of these diverse constituencies to ensure that both nomenclature and diagnostic criteria meet their specific needs. The last thing that is needed is a cross-disciplinary debate on the syndrome's name. Let's not mirror the Middle East, but rather let's establish our highly collaborative two states forthwith.