Loose ends left by the latest PCOS guidelines

Published 12 June 2019

Last year's international guideline for the assessment and management of polycystic ovary syndrome was the first to apply a broad evidence-based approach to improving care, outcome and quality of life in this common condition. Yet despite the evidence, several questions remained unanswered, which a Campus meeting in May sought to address.

Despite its 200 pages, its contribution from numerous experts, organisations and consumers, and north of 70,000 downloads in its first year, the International evidence-based guideline for the assessment and management of Polycystic Ovary Syndrome (PCOS), introduced at ESHRE 2018, has still sufficient gaps in knowledge to warrant a Campus meeting in May 'to tie up the loose ends'.(1,2)

Most of the guideline's 166 recommendations, said the development group's leader Helena Teede from Australia's National Health and Medical Research Council Centre for Research Excellence in PCOS at Monash University, Melbourne, 'are holding up very well', but questions remain over risks and prevention, treatment, long-term cardiovascular effects and quality of life as 'the biggest problem for women'. There was also much equivocal discussion over the symptoms defining the syndrome, their diagnostic applicability, and the resulting phenotypes.

Diagnostic criteria
Overall, however, there seemed to be consensus that the Rotterdam diagnostic criteria - developed by an expert ESHRE and ASRM collaboration in 2003 - are still applicable, with many participants insisting that PCOS is a single syndrome in which the individual patient, not the phenotype, should be treated. The Rotterdam consensus, which today is still Human Reproduction's most cited paper ever, concluded that PCOS is a syndrome of irregular cycles, hyperandrogenism and polycystic ovary morphology - with its diagnosis defined by any two of these three 'cardinal features'.(3) Significantly, the development group for this latest evidence-based guideline 'unanimously supported the Rotterdam diagnostic criteria for adult women' - but questions still remained, not least about diagnosis and prognosis in adolescents, which Rotterdam ignored.

But the first question raised by this Campus meeting (organised in Bilbao by ESHRE's SIG Reproductive Endocrinology) was whether the acceptance of the Rotterdam criteria had increased the prevalence of PCOS. Frank Broekmans' view suggested 'yes', maybe by as much as '50%', which he explained largely by a diagnostic switch from androgen excess to ovarian morphology. Nevertheless, the diagnostic criteria, he said, are likely to get tighter with more sensitive ultrasound scanning (which has already seen the morphological threshold rise from 12 to 20 follicles) and the potential of AMH levels to guide diagnosis. The guidelines, however, recommend that ultrasound is not necessary in subjects with hyperandrogenism and ovulatory dysfunction.

The place of AMH in the diagnosis of PCOS was addressed by Joop Laven from the Erasmus Medical Centre in Rotterdam. He showed data reflecting the well established  association between AMH concentrations and normal ovarian function, and some data reflecting ovarian morphology and PCOS itself, but he was generally sceptical of the studies, noting their heterogeneity in design and outcome (a complaint voiced by many others on different topics at this meeting). As a result Laven concluded that AMH is not yet an acceptable diagnostic alternative for either polycystic ovarian morphology or PCOS itself. However, he did concede that AMH has the potential for detecting ovarian morphology, but not until studies are validated in large populations of different ages and ethnicities.

Indeed, ethnicity (as well as adolescence) was a question brought up by several speakers and one raised by Adam Balen in his presentation on reliably defining anovulation and hyperandrogenism. Balen, who led the WHO's 2016 Guideline on Management of Infertility for PCOS, noted that differing expressions of hyperandrogenism, obesity and insulin resistance are all evident in different ethnic populations and may affect metabolic risk, skin sensitivity to androgens, response to ovarian stimulation and quality of life. Ethnicity, he said, remains a knowledge gap in the latest guideline.

Quality of life was a further knowledge gap in PCOS explored in Bilbao by Terhi Piltonen from Oulu University Hospital in Finland. She noted a substantial prevalence of depression and anxiety in women with PCOS, expressed in eating disorders, body image (hirsutism and weight especially) and psycho-sexual worries. While the new guidelines did recognise these problems, their management was largely ignored, leaving Piltonen to name awareness, screening, early intervention and referral as 'priorities' for their improvement.

Pharmacological treatments
Treatment, which for infertility was covered by a further ESHRE/ASRM consensus in 2007, took on a broader range in the latest guidelines, which noted that 'doctors often focus on individual features of PCOS such as infertility, rather than taking on a broader approach to care'. Terhi Piltonen in reviewing current treatments for symptoms, reaffirmed that 'lifestyle should be first-line for prevention and treatment', with individual patient concerns and quality of life the target for intervention. For managing irregular cycles the pharmacological choice lay mainly between the combined Pill and metformin, with the former described as effective (though metformin was preferred in obese women). For anovulatory infertility Roy Homburg first recommended weight loss followed by clomiphene citrate or the aromatase inhibitor letrozole for the induction of ovulation. While the guidelines do recommend letrozole as first-line treatment for infertility, Homburg did point out that it remains non-indicated in many jurisdictions - despite its better record in live birth and multiple pregnancy rates when compared to clomiphene. Metformin for ovulation induction, either alone or as an adjuvant to clomiphene, was 'not recommended' as first-line treatment by Homburg, who did, however, support low-dose FSH, despite its higher cost and relative lack of evidence.

Second-line therapies for women who fail to ovulate in response to clomiphene or letrozole were not discussed in any detail atthis meeting, but are covered by the new guideline and include low dose gonadotrophins for ovulation induction or laparoscopic ovarian drilling/diathermy. Studies of the nutritional supplement inositol were reviewed by Daniela Romualdi, who concluded that, with insufficient evidence, its place in PCOS should remain 'experimental'.

Stratis Kolibianakis reviewed the final treatment option for PCOS patients, IVF, drawing attention to a deluge of recent data on high responders and the 'segmentation' approach of his former colleagues in Brussels for the prevention of OHSS. While freezing all embryos for later transfer remains controversial in normo-responders in IVF, there is little doubt, said Kolibianakis, that the risk of OHSS is reduced in high responders - as it similarly is by following a GnRH antagonist protocol with a GnRH agonist (and not hCG) as the trigger.

This meeting certainly did not tie up every loose end in PCOS, and one which fluttered loosely throughout this two-day review was the place of obesity (and weight loss) as a cause, symptom or consequence of the condition.(2) The Spanish endocrinologist Hèctor F. Escobar‐Morreale was first asked if metabolic abnormalities, including diabetes and obesity, were 'in or out' as causative features of PCOS, but after reviewing a large body of evidence his opinion was 'definitely out'. The relationship, he argued, was more 'association' than causation.

It was thus salutary after such insistence on lifestyle as the first target of treatment to hear the development group's lead Helena Teede report that 75% of lean PCOS subjects still have high levels of insulin resistance and that 'lifestyle has been largely unsuccessful in normalising BMI'. Thus, added Teede, while a 5-10% weight loss is an achievable goal, more ambitious weight loss targets may have little effect in the larger picture of PCOS. Indeed, said Escobar‐Morreale, a 1 kg loss of weight has no clinical effect'. Nor, he added, are weight-loss drugs an option for severely obese women, where bariatric surgery remains the only proven treatment.

Otherwise, a round-table discussion described the paradox of 'high-quality funding and low-quality research' as the 'biggest problem' in developing the PCOS guidelines. One solution, said Helena Teede, would be in more research collaborations and the prospective study of large cohort of subjects from different ethnic backgrounds.

1. For summary, see Teede HJ, Misso ML, Costello MF, et al. Recommendations from the international evidence-based guideline for the assessment and management of polycystic ovary syndrome. Hum Reprod 2018; 33: 1602-1618.

2. The 'loose ends' discussed at this meeting related to

- Prevalence according to Rotterdam diagnostic criteria

- Phenotypes

- Hyperandrogenism and its definition

- Ovarian morphology as detected by ultrasound

- AMH in diagnosis

- Metabolic abnormalities

- Risk assessment at different life stages

- Quality of life

- Genetic and epigenetic influence

- Lifestyle

- Drugs for symptom relief

- Inositol

- Obesity

- Drug treatments for infertility

- IVF for infertility

3. The Rotterdam ESHRE/ASRM-sponsored PCOS consensus working group. Revised 2003 consensus on diagnostic criteria and long‐term health risks related to polycystic ovary syndrome (PCOS). Hum Reprod 2004; 19: 41-47.

Get notified of new articles with our ESHRE newsletter.

Sign up and never miss an update