Acknowledgments
This work was supported by Grants PI1501686, PIE1600050, and PI1801122 from Instituto de Salud Carlos III, Spanish Ministry of Economy and Competitiveness, and co-financed by the European Development Regional Fund “A way to achieve Europe” (EDRF). CIBERDEM and IRYCIS are also initiatives of from Instituto de Salud Carlos III.
Competing interests statement
The author declares no competing interests.
Introduction
Polycystic ovary syndrome (PCOS) is nothing more, but also nothing less, than a very common female phenotype consisting of a combination of signs and symptoms of androgen excess (hirsutism and/or hyperandrogenemia) and ovarian dysfunction [oligo-ovulation and/or polycystic ovarian morphology (PCOM)], provided that other specific diagnoses such as hyperprolactinemia and nonclassic congenital adrenal hyperplasia have been excluded . The prevalence of PCOS in premenopausal women ranges from ~ 6% (using the older more restrictive criteria) to ~ 20% (when applying current more inclusive definitions) , possibly making this syndrome the most common endocrine and metabolic disorder in women of reproductive age .
Even though PCOS is quite homogeneous in a few characteristics such as its peripubertal onset and similar worldwide prevalence , the PCOS phenotype may result from different combinations of pathophysiological mechanisms making heterogeneity intrinsic to the syndrome . I will here provide hypotheses to explain how the pathophysiological heterogeneity of PCOS is translated into most aspects of the disorder, from clinical presentation to management, but very particularly to its definition.
Etiology and pathophysiology of PCOS
Familial aggregation
Familial aggregation of PCOS and its related traits suggests a genetic basis . However, only the association of a few genetic variants and mutations have been replicated in different populations of patients with PCOS, accounting only for around 10% of the heritability of PCOS . PCOS is now considered a complex multigenic disorder with predisposing and protective genetic variants interacting with strong environmental influences to result in the different PCOS phenotypes . We hypothesized that transmission over generations of these variants may be the result of the survival advantage provided by them. Accordingly, causes of survival advantage in women with PCOS and their progeny include a rapid maturation of the reproductive axis; assertive behavior secondary to increased androgen secretion; subfertility resulting in larger interval between pregnancies, decreasing the birth rate and favoring maternal and infant survival; and oligo-ovulation leading to pregnancies occurring at older ages—when women are more fit than very young women to care for their babies—favoring their survival .
To further contribute to the etiological heterogeneity of PCOS, the environmental factors include diet and lifestyle issues that are heavily influenced by race and/or ethnicity . Accordingly, the genes interfering with these environmental factors might differ depending on the population studied explaining, at least partly, the repeated difficulties in replicating the association of PCOS with genomic variants and loci in populations of diverse origin .
In addition, familial aggregation of PCOS might be related to certain environmental factors that are only present in the affected families and not in the families of unaffected women and might initiate epigenetic mechanisms—that is, stably heritable phenotypes resulting from changes in a chromosome without actual alterations in the DNA sequence . Epigenetic initiators include mechanisms during fetal and childhood development, exposure to environmental chemicals such as endocrine disruptors and certain drugs, the aging process, and diet , most of which might contribute to PCOS .
Androgen excess and insulin resistance: Cause or consequence?
Albeit androgen excess is the main contributor to PCOS , insulin resistance and compensatory hyperinsulinism contribute to androgen excess in these patients . Animal and human studies indicate that insulin acts as a co-gonadotropin on the ovary , facilitates androgen secretion from the adrenal glands and modulates luteinizing hormone pulsatility . Accordingly, the prevalence of PCOS is increased, compared with the general population, in patients with any disorder characterized by systemic hyperinsulinism, either endogenous—obesity , gestational diabetes and type 2 diabetes mellitus , syndromes of extreme insulin resistance caused by mutations in the gene that encodes the insulin receptor or from autoantibodies against the insulin receptor , portosystemic shunts , or even in women with insulinomas —or exogenous such as in type 1 diabetes mellitus . The fact that PCOS is not universal in women presenting with insulin resistance and hyperinsulinism and that insulin resistance is not universal in PCOS suggest that a primary defect that favors androgen excess is essential for the development of the syndrome in response to insulin or other triggering factors .
Also, androgen excess is not only a major mechanism in the oligo-ovulation and cutaneous manifestations of the syndrome, but also facilitates insulin resistance and metabolic dysfunction in women with PCOS by favoring abdominal and visceral adiposity . My colleagues and I hypothesized a decade ago that PCOS results from a vicious circle of androgen excess favoring abdominal adipose tissue deposition and visceral adiposity that, by inducing insulin resistance and compensatory hyperinsulinism, further facilitates androgen secretion by the ovaries and adrenal glands in women with PCOS ( Fig. 1 ) .
The PCOS phenotype covers a spectrum of etiological contributions of a primary defect in androgen secretion mixed with triggering factors
From a pathophysiological perspective, the most important factors responsible for the heterogeneity of PCOS are perhaps obesity, abdominal adiposity, and insulin resistance. Our unifying hypothesis explains PCOS as originated from a primary defect in steroidogenesis leading to androgen excess, the severity of which is quite variable and might be triggered by several factors, among which the best known are those related to excess weight ( Fig. 2 ). The relative contribution of androgen excess and triggering factors to the phenotype of individual patients with PCOS covers a wide spectrum. In one extreme of this spectrum, which may be represented by lean women without any evidence of visceral adiposity or insulin resistance, androgen excess alone is severe enough to cause PCOS without the contribution of any other factors. In the other extreme of the spectrum, PCOS only fully manifests when obesity, abdominal adiposity, insulin resistance, and/or hyperinsulinism trigger a very mild alteration in steroidogenesis. These patients might represent “secondary” forms of PCOS—that will be reviewed in the next chapter—as the PCOS phenotype in these patients frequently resolves once the triggering factor is removed . This spectrum explains the heterogeneity of PCOS symptoms in patients with differing contributions of androgen excess, obesity, and insulin resistance to their individual phenotype ( Fig. 2 ). Obviously, the most severe PCOS phenotypes develop in women with marked obesity or another factor that aggravates a substantial preexisting steroidogenic abnormality. However, some women who exhibit extreme obesity and insulin resistance do not develop PCOS ; therefore, a prerequisite for developing PCOS would be an excess in androgen secretion .
Historical definitions of PCOS: Cause or consequence of heterogeneity?
The term PCOS was introduced after Stein and Leventhal reported in 1935 the combination of hirsutism, amenorrhea, chronic anovulation and infertility, obesity, and enlarged cystic ovaries . Even though they entitled their article as “Amenorrhea and polycystic ovaries,” and version 10 of the World Health Organization International Classification of Diseases (ICD-10) refers to PCOS as “E28.2 Polycystic ovarian syndrome” (with “sclerocystic ovary syndrome” and “Stein-Leventhal syndrome” as synonyms), PCOS can be anything but polycystic!
The “polycystic” appearance of the ovaries frequently found in patients with PCOS is caused by the accumulation of ovarian follicles in different stages of maturation and/or atresia which, by definition, are not actual cysts. The use of this misnomer diverts attention from the actual pathophysiology of the syndrome , the name PCOS is perceived as confusing both by primary health care physicians and patients, and it is not unusual for patients and their families to even express unreasonable concerns about the potential for malignancy of such “cysts” . Accordingly, the 2012 National Institutes of Health Office for Disease Prevention-Sponsored Evidence-Based Methodology Workshop on Polycystic Ovary Syndrome recommended that a new name was needed for PCOS. This proposal was supported by certain professional associations, women’s health organizations, and PCOS patient support groups that supported this proposal to ensure that “an alternative name enhances understanding and recognition of the syndrome and its complex features” . To date, none of the names proposed to such avail have been widely accepted .
Moreover, whether or not “polycystic ovaries” are essential for the definition of PCOS have always been at the core of the endless debate about which is the most proper definition of PCOS . Undoubtedly, the heterogeneous nature of PCOS contributed to such debate. As in the Buddhist and Hindi parable of the blind men and the elephant, the definition of PCOS has usually been biased towards giving more emphasis to the areas of interest of the experts involved: Gynecologists focused on ovarian morphology and dysfunction and infertility, Internists and Endocrinologists tended to highlight the hormonal and metabolic dysfunction characteristic of these patients, Dermatologists cared mostly for hirsutism, acne and alopecia, Pediatricians for premature pubarche, etc. These different points of view made it almost impossible to reach a prolonged consensus among the professionals involved in the study and management of a multifaceted disorder such as PCOS, but also hampered progress in our understanding of PCOS, simply because the patients with PCOS included in some studies were entirely different than those in others precluding confirmation of the results across different populations. In other words, the heterogeneity in the definitions of PCOS across studies ended contributing to the heterogeneity of the syndrome in terms of pathophysiology, clinical presentation and management, and association with other disorders.
Nowadays there are three valid PCOS definitions in use, with only minor differences between them that derive from the ways of combining three individual criteria: hyperandrogenism, ovulatory dysfunction, and PCOM . The 2012 NIH-sponsored workshop favored the more ample definition proposed in 2003 by the European Society of Human Reproduction & Embryology and the American Society of Reproductive Medicine —PCOS is diagnosed when two of the three criteria abovementioned are present after specific etiologies have been excluded—but highlighted the need of providing a precise description of the combination of signs and symptoms that led to such a diagnosis ( Table 1 ) in each case, as a way of minimizing the impact of the heterogeneity in the phenotypes covered by the PCOS definition for research studies .
|
Strategies to overcome the limitations imposed by the heterogeneity of PCOS
Changing the name of the syndrome?
As happens with the debate of diagnostic criteria, the different names proposed to substitute PCOS reflect largely the interpretations and beliefs of the proponents causing conflict with those of others. Experts who defend the role of hyperandrogenism proposed names such as functional ovarian hyperandrogenism , hyperandrogenic-chronic anovulation or female functional hyperandrogenism . Others who consider insulin resistance and metabolic dysfunction as the most important etiopathogenic mechanisms proposed names such as metabolic-reproductive syndrome , syndrome XX , or even prevalent cardiometabolic ovary syndrome, which would also take the abbreviation of “PCOS.” Furthermore, names such as “polyfollicular ovarian syndrome” rely only on an assessment of altered ovarian morphology, which is a criterion that is neither necessary nor sufficient to diagnose the syndrome.
Of note, there are also experts and patient’s associations reluctant about renaming PCOS, mostly because this name has been finally included by the WHO as a disease and is increasingly recognized by the general public as an important health problem. Any change in this terminology might draw all previous efforts in the dissemination of knowledge back to square one . I tend to agree with the latter, but giving Stein and Leventhal due credit for their original observations and continue to name PCOS after them may be an elegant alternative . The term “Stein-Leventhal syndrome” has already been used to name PCOS for decades , and it is well-known to health authorities, patients, and other organizations. Using the names of the persons providing the first description of a disease to term it is not uncommon in modern medicine; i.e., using “Graves Basedow disease” to term autoimmune hyperthyroidism has not caused any problem for close to two centuries .
Changing the definition of PCOS?
Fortunately, after the Evidence-Based Methodology Workshop on Polycystic Ovary Syndrome mentioned earlier , most experts in the field consider the debate closed. If any, in my humble opinion, the ESHRE/ASMR criteria are too ample. Yet this amplitude, which on the one hand may contribute to heterogeneity, on the other hand, has the advantage of covering all phenotypes possibly associated with the term PCOS, including phenotypes lacking any evidence of androgen excess. The current definition reflects perfectly the intrinsic heterogeneity of PCOS but, at the same time, every effort must be made to avoid the frequent belief that a diagnosis of PCOS has the same implications for the present and future health of all patients!
Each individual criteria used to define PCOS has its own clinical consequences. Androgen excess may result in cutaneous manifestations such as hirsutism, acne, and alopecia ; ovulatory dysfunction and chronic oligomenorrhea might result in infertility and endometrial hyperplasia and/or carcinoma ; and isolated PCOM is only associated with a risk of ovarian hyperstimulation syndrome during ovulation induction . As a rule of thumb, the more criteria met by the individual patient with PCOS, the more severe her particular phenotype .
The most severe clinical manifestation is the classic PCOS phenotype that combines hyperandrogenism with oligo-ovulation, regardless of the presence of PCOM. The next phenotype in severity is ovulatory PCOS (presents with hyperandrogenism and PCOM), and the nonhyperandrogenic phenotype, which consists of oligo-ovulation and PCOM, is the milder phenotype ( Fig. 3 A ) .
