Endocrine Feedback Loop: EFL051 – PCOS Subtypes Redux
Host: Dr. Chase Hendrickson (Vanderbilt University Medical Center)
Contributors: Dr. Anna Goldman (Harvard Medical School/Brigham and Women’s Hospital) & Dr. David Ehrman (University of Chicago)
Date: July 18, 2024
Episode Overview
This episode of Endocrine Feedback Loop dives into the evolving classification of polycystic ovary syndrome (PCOS) through the lens of a recent study (forthcoming in Journal of Clinical Endocrinology and Metabolism) led by Kim Vanderham and Lois Mulheitsen. The discussion explores the transformation from traditional "lumping" approaches in PCOS diagnosis toward refined "splitting" into meaningful biological subtypes, using sophisticated clustering analyses. The expert panel reviews the study’s methodology, findings, clinical implications, and limitations, ultimately pondering whether these advances could reshape precision medicine for PCOS.
Key Discussion Points & Insights
Background: PCOS and Diagnostic Criteria
Anna Goldman [02:40]:
- PCOS affects 5–15% of reproductive-aged girls and women.
- Diagnosis is based on two or more of the following: ovulatory dysfunction, hyperandrogenism (clinical/biochemical), and polycystic ovarian morphology.
- Associated with obesity, insulin resistance, and increased risk for cardiovascular disease/diabetes.
David Ehrman [03:32]:
- Pathophysiology centers on altered GnRH pulsatility, leading to increased LH (androgen excess) in the hypothalamic-pituitary-ovarian axis.
- Classic diagnostic criteria:
- NIH (1990/1992): Requires both ovulatory dysfunction and hyperandrogenism.
- Rotterdam: Broader; includes ovarian ultrasound findings (polycystic morphology).
- Ultrasound’s utility is debated; morphological findings are imperfectly correlated with the syndrome.
Quote:
“At least 20% of women without PCOS can be found to have radiographic criteria for polycystic ovaries. And about 20% of women with PCOS will not have polycystic ovaries.” — Dr. Ehrman [08:10]
Current Phenotypes & Genetic Subtyping
Dr. Goldman [09:10]:
- Rotterdam yields four "phenotypes" (A–D), but these do not map cleanly to distinct biology.
- Recent genetic studies suggest three meaningful subtypes (from NIH-defined PCOS):
- Reproductive subtype: High LH & SHBG, low BMI & insulin
- Metabolic subtype: High glucose & insulin, low SHBG & LH
- Background subtype: No clear distinguishing features
Dr. Ehrman [10:25]:
- Clinical significance: Subtypes provide insight for risk stratification (esp. for metabolic complications).
- Insulin/BMI are not diagnostic criteria, but are clinically informative.
Quote:
“What I believe is that the syndrome is misnamed...it should be called the reproductive metabolic disorder...that better reflects the constellation of metabolic and reproductive disturbances.” — Dr. Ehrman [10:30]
Study Purpose & Methods
Host (Dr. Hendrickson) [12:34]:
-
Goals:
- Assess if genetic subtypes are present in Rotterdam-defined PCOS.
- Investigate differences in additional PCOS-related traits across subtypes.
- Explore distinct biological underpinnings by subtype.
-
Design: Cross-sectional, single-center cohort (Erasmus Medical Center, Rotterdam, 1993–2021).
- N = 2,510 (final N = 2,502 after exclusions); ages 13–45; all European ancestry.
- Diagnosis evolved with prevailing criteria (pre-2003 to post-2018 standards).
- Exclusions: recent contraceptive use, non-fasting labs, confounders ruled out.
-
Analysis:
- Unsupervised clustering (using BMI, testosterone, SHBG, DHEAS, LH, FSH, insulin, glucose).
- Subtyped by both Rotterdam and NIH definitions.
Quote:
“What a cluster analysis allows you to do is...identify if there are distinct subgroups based on how the different variables cluster together.” — Dr. Hendrickson [15:21]
Key Results
Dr. Goldman [18:12]:
-
Three subtypes identified across 2,502 PCOS cases:
- Metabolic (41%): High BMI, glucose, insulin, low LH/SHBG
- Reproductive (18%): High LH/SHBG, low BMI/insulin
- Background (41%): No clear biomarker pattern
-
Subtype-specific clinical features:
- Reproductive: Higher AMH and follicle counts.
- Metabolic: Higher triglycerides/LDL, lower HDL, elevated BP, higher DHEAS.
-
Phenotype Correspondence:
- Metabolic group mostly “classic” phenotype A (72%); reproductive subtype had more mixed A/D; background predominantly D.
-
Cluster prevalence shifts with diagnostic criteria used:
- Classic NIH-defined PCOS: Metabolic subtype much more prevalent (62%)
- Non-NIH Rotterdam-defined PCOS: Background subtype dominates
Notable Moment:
“The biomarkers they used were consistent with those clusters. So it’s not diagnostic...but it’s consistent with the reproductive versus metabolic phenotype.” — Dr. Ehrman [20:13]
Discussion & Clinical Implications
Dr. Hendrickson [22:29]:
-
PCOS is highly heterogeneous; classic criteria may obscure underlying biology.
-
Genetic subtypes “map” onto distinct biological pathways—findings now reproduced in a European/Dutch cohort.
-
Key clinical takeaways:
- Metabolic subtype faces higher cardiovascular and metabolic risk.
- Reproductive subtype uniquely characterized by abnormal folliculogenesis markers (higher LH, SHBG, AMH).
- No difference in age across clusters—features emerge young.
Quote:
“It may be that the metabolic aspects of the disorder are already present from an early age...the surveillance for these metabolic disturbances should probably be made more frequently or more intensively.” — Dr. Ehrman [25:04]
Strengths & Limitations
Strengths (from authors):
- Large, thoroughly phenotyped, single-center cohort
- Extensive trait exploration beyond diagnostic criteria
- Mapping of clusters onto existing phenotypic guidelines
Limitations:
- Referral Bias: As a single-center specialty clinic, may not reflect all patients.
- Retrospective, cross-sectional design: No longitudinal outcomes.
- European ancestry only: Limited generalizability; need for diverse cohorts.
- No validation with uncorrelated biomarkers.
Quote:
“You can’t really extend the findings to other populations. But I think those data are forthcoming.” — Dr. Ehrman [28:52]
Memorable Quotes & Takeaways
Precision Medicine Potential
- “Clustering enables data driven diagnosis of PCOS and the identification of these subtypes will allow for precision medicine approaches.” — Study authors, summarized by Dr. Hendrickson [29:03]
Practical Clinical Advice
- “It’s important to make sure that they meet at least the NIH criteria and/or the Rotterdam criteria before making that diagnosis. I think it does a disservice to over diagnose this and also to under diagnose it.” — Dr. Ehrman [31:12]
Expert Appraisal & Future Directions
Dr. Goldman [29:43]:
- Study helps clinicians better explain the spectrum of PCOS to patients—not a monolithic diagnosis.
- Improved subtyping may inform more precise treatment options (e.g., tailoring the “triad” of OCPs, metformin, antiandrogens, inositols).
Dr. Ehrman [31:12]:
- Subphenotyping is promising but not yet actionable in routine clinical practice.
- Need for replication and further studies, but clinicians should be vigilant about early metabolic risk and consider more intensive screening for at-risk subtypes.
Recommended Timestamps by Topic and Speaker
- [02:40]: PCOS diagnosis and pathogenesis overview (Goldman/Ehrman)
- [09:10]: Existing phenotypes vs. new genetic subtypes (Goldman)
- [12:34]: Study purpose and methods (Hendrickson)
- [18:12]: Core results: Subtype definitions & prevalence (Goldman)
- [22:29]: Discussion: Clinical meaning and translation (Hendrickson/Ehrman)
- [26:58]: Strengths & limitations (Hendrickson/Ehrman)
- [29:43]: Clinical appraisal of study (Goldman)
- [31:12]: Practicality of subtyping in care (Ehrman)
Final Thoughts
This episode illustrates the growing sophistication in PCOS research, as clustering analyses begin to “split” what was formerly “lumped” under a broad syndrome label. For clinicians, the study affirms the underlying biological complexity of PCOS, highlights the importance of differentiating patients with prominent metabolic risk, and hints at a future where precision medicine can personalize care for women with PCOS.