The core insight of the model is that the same metabolic and endocrine abnormalities, insulin resistance, androgen excess, and chronic low-grade inflammation, can produce very different tissue states depending on how long they persist and how the affected cells respond. In the earliest, plasticity-dominant stage, tissues retain their adaptive capacity despite the presence of metabolic and hormonal disturbances. Cells can still compensate, signaling networks remain flexible, and the pathological changes are potentially reversible. This stage corresponds to what clinicians often observe in younger patients, whose phenotypic characteristics can shift over time and who may even regain regular menstrual cycles as they age, a phenomenon documented in long-term follow-up studies.

The transition to the second stage hinges on a failure of cellular housekeeping. At the center of this failure sits the mechanistic target of rapamycin, or mTOR, a signaling hub that integrates nutrient, growth factor, and stress cues. In PMOS, chronic hyperinsulinemia and androgen excess activate mTOR complex 1, which in turn suppresses autophagy, the recycling process by which cells clear damaged proteins and organelles. Studies in mouse models of the syndrome have shown that dehydroepiandrosterone-induced activation of mTORC1 and inhibition of autophagy contribute to skeletal muscle insulin resistance, while work on granulosa cells, the support cells surrounding the oocyte, has linked PI3K/AKT/mTOR pathway dysregulation to impaired autophagy and apoptosis. When autophagy falters, damaged mitochondria accumulate, and mitochondrial dysfunction becomes a self-reinforcing problem.

Mitochondrial impairment is more than a downstream casualty in this framework; it is an engine of progression. Dysfunctional mitochondria generate excess reactive oxygen species, which oxidize mitochondrial DNA and further degrade respiratory chain function. Reviews of mitochondrial DNA mutations and dysfunction in the syndrome, along with measurements of mitochondrial DNA oxidation and copy number across different metabolic phenotypes of affected women, point to a consistent pattern of energetic failure. In granulosa cells, restricted glycolysis and reduced energy production have been identified as a dominant problem, and abnormal energy metabolism in these cells correlates with poorer outcomes in in vitro fertilization-embryo transfer. NAD+ deficiency in the granulosa cells of affected women compounds the deficit, weakening sirtuin-dependent repair pathways and accelerating the slide toward the third stage.

That third stage, the pathological-remodeling state, is where the model makes its most provocative claim. Here, cellular senescence takes center stage. Senescent cells, which permanently exit the cell cycle under the control of tumor suppressor programs involving p16INK4a and p21Cip1/Waf1, do not simply sit quietly. They secrete a cocktail of pro-inflammatory cytokines, growth factors, and matrix-remodeling enzymes known as the senescence-associated secretory phenotype, or SASP. Recent work has demonstrated that granulosa cells in the syndrome exhibit markers of senescence, and RNA sequencing of these cells has revealed coordinated disruptions in senescence and sphingolipid metabolism. The SASP-driven inflammation feeds back into insulin resistance, creating a vicious cycle in which metabolic dysfunction, inflammatory signaling, and cellular aging amplify one another.

The structural consequences of this cycle are tangible. Fibrosis, the excessive deposition of extracellular matrix that stiffens and scars tissue, has emerged as a key feature of the remodeled ovary. A striking recent study showed that deficiency of the antioxidant enzyme glutathione peroxidase 4 induces ferroptosis, an iron-dependent form of cell death, which in turn drives endometrial epithelial fibrosis in the syndrome. Ovarian fibrosis itself reduces tissue plasticity, impairing follicular development and angiogenesis, and endometrial stromal fibroblasts from affected women display impaired progesterone-mediated decidualization, aberrant cytokine profiles, and enhanced immune cell migration, all of which undermine implantation and pregnancy success. In this state, the disease becomes self-sustaining: the tissue can no longer adapt, and the pathological architecture maintains the disorder even if the original metabolic triggers are partially corrected.

What makes the three-stage model scientifically valuable is that it integrates previously parallel lines of evidence into a single mechanistic narrative. The insulin resistance story, the hyperandrogenism story, the inflammation story, and the senescence story are no longer competing explanations but sequential and overlapping layers of the same process. Insulin resistance, documented in affected women through euglycemic-hyperinsulinaemic clamp studies, drives compensatory hyperinsulinemia, which stimulates ovarian androgen production both directly and by suppressing hepatic sex hormone-binding globulin. Elevated androgens and insulin-like growth factor 1 signaling converge on mTORC1, suppressing autophagy and promoting theca cell expression of CYP17A1, the key androgen-synthesizing enzyme, via ROS-dependent p38 and JNK signaling when autophagy is blocked. Each loop tightens the next.

The model also offers a biological explanation for the heterogeneity that has frustrated clinicians for decades. Data-driven clustering studies have identified subtypes of the syndrome with different phenotypic characteristics and clinical outcomes, and genome-wide association signals at loci such as DENND1A, THADA, and genes encoding steroidogenic enzymes underscore the contribution of inherited susceptibility. In the staging framework, genetic and epigenetic vulnerability, including transgenerational epigenetic transmission demonstrated in animal models, determines how quickly an individual progresses along the spectrum, while environmental factors such as diet, gut microbiota composition, and micronutrient status modulate the pace. A lean woman with irregular cycles and preserved metabolic flexibility may sit in the plasticity-dominant stage, whereas an older patient with metabolic syndrome, fatty liver, and ovarian fibrosis may have reached the remodeling stage, even if both satisfy the same Rotterdam criteria.

Therapeutically, the implications are significant, because different stages may demand different interventions. In the plasticity-dominant stage, lifestyle modification, exercise, and insulin-sensitizing drugs such as metformin, which has been shown in rat models to ameliorate the syndrome by normalizing excessive autophagy in granulosa cells through the PI3K/AKT/mTOR pathway, could restore adaptive capacity before irreversible damage occurs. GLP-1 receptor agonists, berberine combinations acting on the AMPK/AKT/mTOR axis, and agents targeting ferroptosis or senescence may be most valuable at the transition point, when autophagy is failing but fibrosis has not yet taken hold. Once the pathological-remodeling state is established, senolytic or senomorphic approaches, antifibrotic strategies, and regenerative interventions such as mesenchymal stem cell transplantation, which has improved ovarian mitochondrial function in animal models via the PI3K-AKT pathway, or exosome-based therapies, become the more plausible options. Rapamycin treatment in mouse models of the syndrome has already shown effects on follicle development, hinting that mTOR modulation could be staged therapeutically.

The author is careful to position the model as complementary to, rather than a replacement for, existing diagnostic systems, and as a narrative review it synthesizes rather than tests the framework experimentally. Validation will require longitudinal studies that track molecular markers of autophagy, mitochondrial function, senescence, and fibrosis in patients over time, and the development of biomarkers capable of assigning individuals to a stage. Yet the conceptual shift is already compelling: PMOS is not a fixed diagnosis but a trajectory, and recognizing where a patient stands on that trajectory, from flexible and reversible to scarred and self-sustaining, could transform how clinicians predict prognosis, sequence treatments, and ultimately prevent the metabolic and reproductive consequences of one of medicine’s most heterogeneous disorders.

Subject of Research: A three-stage biological disease-staging model of polyendocrine metabolic ovarian syndrome based on signaling dysregulation, autophagy, mitochondrial dysfunction, and cellular senescence

Article Title: A Biological Disease-Staging Model of Polyendocrine Metabolic Ovarian Syndrome: From Plasticity-Dominant States to Pathological Remodeling

Article References: Kobayashi, H. (2026). A Biological Disease-Staging Model of Polyendocrine Metabolic Ovarian Syndrome: From Plasticity-Dominant States to Pathological Remodeling. Reproductive Sciences. https://doi.org/10.1007/s43032-026-02191-8

Image Credits: AI Generated

DOI: 10.1007/s43032-026-02191-8

Keywords: polyendocrine metabolic ovarian syndrome, PCOS, insulin resistance, hyperandrogenism, mTOR signaling, autophagy, mitochondrial dysfunction, cellular senescence, SASP, ovarian fibrosis, Rotterdam criteria, disease staging