Recently identified form of PCOS has implications on fertility treatment strategies and outcomes
PCOS is a widely known and extensively studied medical diagnosis, closely associated with female infertility but also later in life with the so-called metabolic syndrome, characterized by significantly increased risks for type II diabetes mellitus, hypertension and atherosclerotic heart disease. It, however, also is a widely misunderstood diagnosis, mostly because it, in reality, does not represent one single condition but a basket of conditions, in principle (though even in that not universally) characterized by what is called a PCOS-ovarian-phenotype of increased ovarian volume and pearl chain-like follicles on ultrasound.
Over decades, various professional organizations have attempted to classify the individual phenotypes of PCOS but, not only was consensus never really achieved, but the last classifications (i.e., the so-called Rotterdam Criteria) are already totally outdated since they do not even include what, likely, represents the most typical single finding in all PCOS phenotypes: very abnormally high anti-Müllerian hormone (AMH) in peripheral blood.
By many considered the most frequent cause of female infertility, the reason why PCOS causes infertility has historically been attributed to anovulation (failure to ovulate). Indeed, one of the two most frequently observed phenotypes of PCOS, the so-called classical phenotype, is characterized by anovulation. This phenotype is also characterized by truncal obesity and hyperandrogenism (high testosterone levels) with peripheral manifestation including hirsutism (excessive hair growth). It is also this PCOS phenotype that is associated with the metabolic syndrome. Because affected women by this PCOS phenotype are visually very easy to recognize, the classical phenotype in many ways has become the general “image” of PCOS. This is, however, yet another reason why PCOS, even among many medical experts, is so widely misunderstood.
Classical PCOS represents only ca. 40% of all PCOS cases. Approximately another 40% of PCOS involves the so-called lean phenotype. These are women who do not represent any of the peripheral and/or laboratory findings described above for women with classical PCOS, except for the fact that both phenotypes are characterized by abnormally high AMH levels. Lean PCOS patients, however, have normal and, often, even low weight, may be hyperandrogenic but usually do not demonstrate signs of hirsutism and do not demonstrate increased risk for the metabolic syndrome. They also often present with regular menses and are, therefore, not anovulatory; yet often, still, do not conceive. Though they represent approximately equal numbers to classical phenotypes, the perception of PCOS as an anovulatory condition, nevertheless, persists not only among the lay pubic but also within the medical community. In lean PCOS patients, the cause of infertility, therefore, cannot be anovulation and, as we will discuss below, was now largely elucidated by CHR investigators.
Approximately another 10-20% of PCOS patients under Rotterdam Criteria represent a third phenotype, often referred to as non-PCO PCOS (i.e., not showing the usual pearl chain-like pattern on ultrasound).
Because of the dichotomy in presentation between classical and lean phenotypes, classical PCOS patients for the longest time were widely considered the “more severe” and “poorer prognosis” patients. As we will demonstrate below, that, too, has been a misunderstanding that has greatly contributed to much of the confusion surrounding PCOS. While classical PCOS patients because of their risk of developing metabolic syndrome in a medical sense, indeed, must be considered “higher risk” patients, from a fertility stand point, the lean-PCOS patient has proven much more resistant to fertility treatments than the classical PCOS patient.
The hypo-androgenic PCOS-like phenotype
Because of much of the confusion surrounding PCOS, CHR’s interest in PCOS research has been longstanding. It, however, was strongly reinvigorated in 2014, when Vitaly A. Kushnir, MD, CHR’s Director of Fertility Preservation and of Continuing Medical Education, initiated a study that attempted to evaluate the ontogeny (evolution over time) of PCOS as women get older. Aside from expected results the study, published a year later (Kushnir et al., J Ovarian Res 2015;8:45), demonstrated a totally unexpected finding in that practically all PCOS patients in CHR’s anonymized electronic research data bank turned out to be lean PCOS patients. Almost none were of the classical PCOS phenotype.
Trying to understand this unexpected finding, CHR investigators found only one possible explanation: Classical PCOS patients, very obviously, conceived before reaching CHR, while lean PCOS patients, after repeatedly failing treatments elsewhere, did finally reach CHR, widely considered a leading IVF center of last resort, at much later stages of their fertility treatment journey. Kushnir’s study, therefore provided the first hint that our understanding of classical and lean PCOS patients required reevaluation. This reevaluation took place under the leadership of Norbert Gleicher, MD, CHR’s Medical Director and Chief Scientist, and led to what, likely, is an almost revolutionary new understanding of PCOS or, at least, of lean PCOS. The findings of these further studies were published in prestigious medical journals in 2017 and 2018, respectively.
The findings, indeed, were in many ways stunning: First, the additional studies demonstrated beyond reasonable doubt that many lean PCOS patients, up to that moment believed to be the “better” and “easier to treat” infertile PCOS patients in comparison to classical PCOS patients, indeed, at least when it came to infertility, were the more difficult to successfully treat. And one of the main reasons was that, in contrast to classical PCOS patients, the cause of their infertility was not anovulation.
What was then the cause of infertility in these lean PCOS patients?
Here is where the story is getting really interesting: As noted before, PCOS patients are usually hyperandrogenic (i.e., demonstrate abnormally high testosterone). Yet, to everybody’s surprise these lean PCOS patients at CHR turned out to almost uniformly demonstrate low testosterone (hypoandrogenic) and, as one would expect in compensation, high sex hormone-binding globulin (SHBG). This surprising finding led to the acronym H-PCOS and raised the question, why were testosterone levels low in these women?
A woman’s testosterone is approximately half and half produced by adrenals and ovaries. Where over- or under-production takes place can be relatively easily evaluated by measuring in parallel dehydroepiandrosterone-sulfate (DHEAS) levels, which is only produced in adrenals. A woman with low testosterone and low DHEAS, therefore, likely has low testosterone due to insufficient adrenal androgen production. And this was exactly the picture that evolved in further studies.
Why did these women suffer from adrenal underproduction of androgen hormones?
Once again, the answer became quickly apparent: Adrenal glands are made up of three distinct layers, called zonae, with each producing a different group of hormones. The zona reticularis is the most inner one and produces androgens. Adrenal insufficiency (Addison’s disease) is a rare autoimmune disease that primarily involves the other two adrenal zonae, shutting down production of so-called glucocorticoids and mineralocorticoids. It, therefore, appeared likely that, if insufficiency of the other two zonae was autoimmune, insufficiency of the zona reticularis, likely, also must be autoimmune in etiology.
This thinking was further supported by the observation that these women to an extremely high degree (almost universally indeed) demonstrated other evidence of autoimmunity. Over 40% demonstrated anti-thyroid autoimmunity alone. CHR investigators, therefore, concluded that the observed hypoandrogenism observed in these women with great likelihood was adrenal in origin and autoimmune in etiology. They, therefore, recommended that insufficiency of the zona reticularis be integrated into the diagnosis of adrenal insufficiency (Gleicher et al., J Clin Endocrinol Metab 2017;102(9)3569-3570).
In further investigating the ontogeny of this condition CHR investigators made additional fascinating observations: At least some (and possibly all) women with lean PCOS at young ages are hyperandrogenic and usually, nevertheless, regularly ovulatory. In late 20s to mid-30s their androgen levels suddenly and surprisingly quickly plunge, while their AMH levels remain elevated. They then go through a period of demonstrating normal androgen levels (from previously high ones) before dropping further into hypo-androgenic levels. Since patients present to CHR usually at older ages, it was not surprising that the overwhelming majority of patients with lean PCOS were already hypo-androgenic.
How is a diagnosis of H-PCOS reached?
The diagnosis of a H-PCOS-like phenotype is reached based on the following characteristics:
- Female infertility
- Lean BMI
- Unusually high AMH for age and/or in proportion to FSH
- Low testosterone, DHEA, DHEAS
- High SHBG
- Evidence of autoimmunity
- Often multiple failed IVF cycles
- Larger than expected egg numbers for age but poor egg quality
- Poor embryo quantity and quality
Why is a timely diagnosis of H-PCOS so important?
It is now more than a decade since the importance of good testosterone levels for normal follicle maturation has been widely recognized. CHR researchers, of course, have been on the forefront of this development. Obtaining good intra-ovarian testosterone levels before IVF via DHEA or direct testosterone supplementation is, however, even more important in H-PCOS patients because their ovarian androgen receptors still carry memory of abnormally high testosterone levels in their youth. The physiological perception of hypo-androgenism is, therefore, in such patients even more pronounced. Androgen pre-supplementation with DHEA or testosterone directly prior to IVF is crucial.
Even if clinically recognized as PCOS patients, most colleagues would never suspect such patients of being hypo-androgenic. Most IVF centers, indeed, do not even evaluate androgen levels in infertile women. Such evaluations are, however, important not only for diagnostic purposes but also for purposes of being able to offer correct prognostic assessments. If a patient’s hypo-androgenism is found to be adrenal in origin, her prognosis immediately improves because, in such a case, androgen supplementation will radically improve her prognostic outlook. If, on the other hand, her androgen deficiency turns out to be ovarian (low testosterone but normal DHEAS), then her prognosis remains poor because such a finding suggests that her ovaries are really “burned out.”
We hope the information provided here helps patients and colleagues alike. The prevalence of H-PCOS in CHR’s patient population, once the clinical presentation became clear to everybody, was truly shocking. This may partially be due to the fact that CHR sits on the top of the referral pyramid, seeing most patients only very late in their respective fertility journeys. CHR may, therefore, see a disproportionate concentration of H-PCOS patients. The diagnosis can, however, be reached much earlier and, now that the phenotype is very clearly elucidated, should be made earlier. As noted above, H-PCOS patients drop their androgens usually starting in their 20s, and treatments can and should be adjusted. Who searches will find!
This is a part of the June 2018 CHR VOICE.