Endometriosis from initiation to innovation: Immune microenvironments, clonal lesion biology, and the emerging therapeutic pipeline

Despite increased awareness and updated guidance, endometriosis remains defined by prolonged diagnostic delay, incomplete symptom control, and a lack of validated disease-modifying endpoints. Patient-centered studies continue to show that delays arise from both patient-side and health-system factors, including normalization of severe dysmenorrhea, fragmented referral pathways, inconsistent imaging access, and variable specialist expertise [ 111 ]. This diagnostic latency is not merely inconvenient: it amplifies financial toxicity, delays fertility planning, and may allow progression from early inflammatory lesions toward fibrotic, neuroangiogenic, and, in ovarian disease, potentially cancer-relevant microenvironmental states [ 111 – 116 ]. Contemporary biomarker reviews underscore that no single circulating marker or biosignature is yet accurate and reproducible enough to replace current diagnostic pathways [ 112 ]. Menstrual effluent is increasingly viewed as a promising non-invasive biospecimen because it is biologically proximal to disease processes, and recent proof-of-concept studies using menstrual-blood inflammatory assays and single-cell approaches support feasibility, but not yet clinical readiness [ 113 , 114 ]. Imaging continues to improve detection of ovarian endometrioma and deep infiltrating endometriosis, especially in expert hands, yet sensitivity remains limited for superficial peritoneal disease and some anatomically complex sites [ 115 , 116 ]. Current therapy also remains constrained by a mismatch between what patients need and what trials usually measure. Most approved medical and surgical treatments are effective primarily for symptom suppression rather than durable disease modification [ 117 – 120 ]. Reviews of hormonal management consistently emphasize that pain reduction must be balanced against hypoestrogenic adverse effects, mood symptoms, irregular bleeding, bone health, and reproductive goals, making shared decision-making essential rather than optional [ 117 , 119 ]. At the same time, recurrence after surgery-particularly after ovarian endometrioma treatment-remains clinically meaningful, underscoring that lesion persistence is driven by biological programs not fully addressed by short-term endocrine suppression [ 118 ]. A key unmet need, therefore, is an accepted framework for defining and measuring disease modification, including lesion regression, prevention of fibrosis and neuroinflammation, reduced need for repeat surgery, and protection of ovarian function. From a translational perspective, this issue also intersects with cancer biology: in ovarian disease, biomarkers that distinguish quiescent lesions from those with persistent oxidative, fibrotic, or clonal progression features could eventually support both treatment stratification and surveillance-oriented research [ 117 – 120 ]. Fertility remains another domain in which care is often reactive rather than strategic. Systematic reviews show that endometriosis does not uniformly reduce assisted-reproduction success across all patient groups, but implantation-related outcomes and ovarian-reserve considerations remain clinically important in selected phenotypes [ 121 ]. The role of surgery before IVF is especially controversial. Contemporary reviews and meta-analyses largely agree that routine excision of ovarian endometriomas before ART does not reliably improve reproductive outcomes and may reduce ovarian reserve, particularly in bilateral or repeat procedures [ 122 , 123 ]. These observations support a more individualized approach in which pain burden, lesion anatomy, retrieval feasibility, ovarian reserve, and suspicion for atypia or malignant transformation are weighed together rather than defaulting to surgery-first algorithms [ 121 – 123 ]. This is another setting in which a mild cancer-angle is relevant: although most endometriomas are benign, cyst characteristics and imaging context can influence whether a conservative fertility strategy is appropriate or whether additional oncologic caution is needed. The under-addressed burden of endometriosis extends well beyond pelvic pain and infertility. Systematic evidence confirms substantial psychiatric morbidity, especially depressive and anxiety symptoms, with distress amplified by pain severity, diagnostic delay, disrupted schooling and employment, and uncertainty about fertility [ 124 ]. Genetic studies also indicate shared biological architecture between endometriosis and gastrointestinal disorders such as irritable bowel syndrome, reinforcing the need for integrated rather than siloed pelvic-pain care [ 125 ]. At a population level, disparities in diagnosis, trial representation, and treatment are increasingly evident. Race and ethnicity remain inconsistently reported in endometriosis research, limiting the development of equitable evidence-based care, while pharmacoepidemiologic analyses suggest real differences in prescribing patterns across racial groups [ 126 , 127 ]. These disparities are not peripheral issues; they directly shape who enters trials, whose biology is reflected in biomarker studies, and which patients ultimately benefit from therapeutic innovation. Methodologic limitations in clinical trials continue to slow translation. Registry-based analyses show that the global trial landscape remains dominated by shorter-duration studies, uneven geographic representation, variable phenotyping, and heavy reliance on pain endpoints over more durable biological or functional outcomes [ 128 , 129 ]. Emerging real-world phenotyping frameworks argue that future studies should integrate surgical data, tissue omics, longitudinal outcomes, digital symptom tracking, and imaging to better capture disease heterogeneity [ 130 ]. Machine-learning protocols and large multi-ancestry omics efforts further support the move toward precision medicine, where diagnosis and treatment selection are informed not only by symptoms and lesion location, but also by immune states, stromal programs, genetic architecture, and possibly ovarian cancer–adjacent risk biology in selected lesion subtypes [ 131 , 132 ]. In other words, the next-generation trial should not simply ask whether a drug reduces pain over a few months; it should ask which biological subtype responds, whether lesion ecology changes, and whether downstream sequelae such as fibrosis, infertility, or cancer-relevant progression markers are altered. Against this backdrop, the endometriosis pipeline has become more diverse over the past several years. Newer reviews of therapeutic development emphasize that the field is moving beyond a purely hormone-centered paradigm toward a broader portfolio that includes immune, inflammatory, fibrotic, angiogenic, neuroimmune, and delivery-focused strategies [ 133 – 145 ]. Oral GnRH antagonists remain the most mature next-generation hormonal platform, with accumulating meta-analytic and narrative-review evidence supporting efficacy for dysmenorrhea and non-menstrual pelvic pain, while also clarifying tradeoffs in bone density, lipid effects, and tolerability [ 136 – 141 ]. The inclusion of add-back strategies, particularly in relugolix-based combination regimens, represents a pragmatic effort to widen the therapeutic window without losing efficacy [ 140 , 141 ]. At the same time, selective progesterone receptor modulators remain of interest but continue to be constrained by safety considerations and the need for longer-term data; vilaprisan provides proof of concept that this axis remains biologically attractive even if clinically unsettled [ 142 ]. The most conceptually important evolution, however, lies in the non-hormonal and advanced-modality space. Recent reviews now frame endometriosis drug development in terms familiar to oncology and chronic inflammatory disease: targetable stromal-immune niches, oxidative-stress circuits, fibrosis pathways, and biomarker-linked patient stratification [ 133 – 135 ]. This is where the mild cancer angle is most scientifically useful. Ovarian endometriosis, in particular, shares enough microenvironmental features with premalignant or tumor-adjacent biology-persistent inflammation, altered angiogenesis, oxidative stress, immune tolerance, and clonal selection-that disease-interception logic borrowed from cancer research may become increasingly relevant. Current pipeline discussions therefore include biologics, antibody platforms, RNA-and gene-based concepts, repurposed drugs, and local delivery strategies alongside endocrine agents [ 134 , 135 ]. Early human or preclinical signals are especially notable for immune-targeted approaches such as IL-1 blockade and IL-8 axis inhibition, which aim to move the field toward lesion-modifying rather than merely symptom-suppressive therapy [ 143 , 144 ]. More broadly, the translational challenge is no longer a lack of ideas, but rather the need to match mechanism, model, biomarker, and patient phenotype in a way that can support durable and fertility-sparing benefit.

Endometriosis is fundamentally an immunopathologic disease in which ectopic endometrial cells are not effectively cleared but instead persist within an immune environment that is chronically activated, incompletely cytotoxic, and biased toward tissue repair, fibrosis, and neuroangiogenesis [ 76 , 77 , 92 , 105 , 106 ]. Rather than functioning as a protective surveillance system, the immune network in endometriosis becomes permissive: innate immune cells are reprogrammed toward lesion-supportive states, adaptive immunity is skewed toward chronic antigen-driven inflammation and tolerance, and neuroimmune circuitry amplifies pain and structural remodeling [ 76 , 77 , 92 ]. This maladaptive immune ecology is now one of the strongest conceptual bases for moving beyond purely endocrine therapies toward microenvironment- and immune-directed intervention strategies [ 105 , 106 ]. Innate immune dysfunction is a defining feature of lesion establishment and chronic persistence. Macrophages are the dominant immune population in peritoneal fluid and lesions, yet in endometriosis they appear to be numerically expanded but functionally impaired [ 77 , 92 , 93 ]. Contemporary single-cell studies identify distinct lesion-associated macrophage phenotypes with prodisease and proresolving functions, including matrix-remodeling, pro-angiogenic, and fibrosis-linked states that differ across lesion locations and likely reflect local niche conditioning [ 93 ]. Macrophage behavior is further shaped by extracellular vesicle signaling: lesion- or uterine-derived exosomes can drive M2-like polarization, reduce phagocytosis, and support lesion development through defined microRNA cargo and receptor-mediated programs [ 85 , 107 – 109 ]. In parallel, annexin A2 suppression and iron/heme-associated stress responses impair macrophage phagocytic capacity, reinforcing the failure of ectopic cell clearance [ 87 , 110 ]. Natural killer (NK) cells also display reduced cytotoxicity in endometriosis. Recent reviews specifically focused on NK-cell dysfunction emphasize impaired effector function and defective elimination of aberrant endometrial targets, thereby enabling immune escape of refluxed or implanted tissue [ 94 ]. Neutrophils contribute primarily to early lesion establishment, where they generate pro-inflammatory and pro-angiogenic conditions favorable to implantation and growth [ 95 ]. Neutrophil extracellular traps have additionally been linked to inflammatory burden and symptom severity, suggesting that NET biology may be relevant not only to pathogenesis but also to biomarker development [ 96 ]. Mast cells occupy a strategic position at the interface of inflammation and nociception. Lesional mast-cell activation through MRGPRX2–histamine/HRH1/TRPV1 signaling promotes sensory-neuron sensitization and pain, while histamine-pathway profiling across lesion subtypes indicates that these programs are broadly integrated into endometriosis-associated inflammatory remodeling [ 97 , 98 ]. Together, innate immune dysfunction in endometriosis is less a matter of excessive immunity per se than of misdirected immunity-an inflammatory response that fails to eliminate ectopic tissue and instead facilitates lesion survival, vascularization, fibrosis, and pain. Adaptive immunity is likewise altered in ways that support chronic disease rather than resolution. T-cell profiles in endometriosis show increased inflammatory skewing, particularly toward IL-17-related programs, alongside changes in regulatory lymphocyte populations that may sustain local immune tolerance [ 99 , 100 ]. Th17-associated inflammation is of particular interest because it links adaptive immunity to neutrophil recruitment, fibro-inflammatory remodeling, and angiogenesis [ 99 ]. At the same time, dysregulation of IL-24 producing regulatory T and B lymphocytes suggests that not all “regulatory” programs are uniformly protective; some may reinforce an immune environment that tolerates persistent lesions [ 100 ]. B-cell involvement has become increasingly evident with recognition of tertiary lymphoid structures and local humoral activation within lesions [ 101 ]. Organized immune aggregates consistent with tertiary lymphoid structures provide a morphological basis for sustained local antigen presentation and adaptive activation. In parallel, autoantibody profiling of peritoneal fluid has identified p53 and other reactivities in subsets of patients, supporting a model of localized humoral dysregulation and immune complexity beyond simple inflammation [ 102 ]. These observations suggest that adaptive immune dysfunction in endometriosis should not be reduced to a single pathway; rather, it reflects persistent, lesion-contextual antigenic stimulation that variably engages inflammatory, regulatory, and autoimmune-like programs. A major advance in the field is recognition that immune dysfunction in endometriosis cannot be separated from neural biology. Recent mechanistic studies show that nociceptor-to-macrophage communication through CGRP/RAMP1 signaling promotes both lesion growth and endometriosis-associated pain, directly linking local immune activation to peripheral sensitization [ 103 ]. Complementary frameworks implicate neuroinflammation, JAK/STAT signaling, mast-cell activation, and histamine-associated pathways in sustaining chronic pain states even when endocrine suppression reduces cyclic activity [ 92 , 97 , 98 , 104 ]. This has major translational implications: it helps explain why pain can persist despite lesion suppression, and why therapies that only reduce estrogen may fail to fully address disease burden in patients with neuroimmune or fibrotic-predominant phenotypes. From a therapeutic perspective, the immunopathology of endometriosis expands the target space far beyond classic estrogen/progesterone pathways. Current immune-focused reviews highlight macrophage programming, cytokine and chemokine signaling, neuroimmune mediators, angiogenesis, and fibrosis-associated pathways as rational targets for next-generation therapy [ 105 , 106 ]. Importantly, these strategies may be especially relevant for patients seeking fertility-sparing and non-contraceptive options, as they offer the possibility of modifying lesion-supportive microenvironments without inducing profound systemic hypoestrogenism [ 91 , 103 – 106 ]. Future clinical progress will likely depend on integrating lesion phenotype, immune signature, and pain mechanism into trial design, rather than treating endometriosis as a uniform endocrine disorder.

Endometriotic lesions are sustained within highly specialized tissue niches rather than existing as isolated ectopic implants. Across peritoneal, ovarian, and deep infiltrating phenotypes, lesion survival depends on reciprocal interactions among ectopic epithelial and stromal cells, mesothelial and fibroblast populations, immune cells, extracellular matrix (ECM), soluble mediators, and neurovascular elements [ 76 – 82 ]. Single-cell and spatial studies now show that lesions occupy remodeled microenvironments enriched for activated stromal states, immunotolerant programs, angiogenic signaling, and matrix-remodeling pathways, supporting the concept that endometriosis is fundamentally a disease of pathological niche formation rather than simple ectopic tissue deposition [ 78 – 83 ]. The pelvic peritoneum is the principal site of superficial lesion implantation and functions as an active biological interface rather than a passive surface. Mesothelial cells, submesothelial fibroblasts, resident immune populations, and peritoneal fluid together create a microenvironment that determines whether refluxed endometrial material is cleared or retained [ 76 , 77 ]. Early lesion formation appears to involve dynamic crosstalk between invading endometrial stromal cells and mesothelial cells, with experimental models showing that the properties of the invading stromal “seed” may be more determinant than the mesothelial “soil” in driving attachment and invasion [ 90 ]. Once implantation is established, single-cell and spatial analyses indicate that mesothelial and stromal compartments are reprogrammed toward pro-adhesive, pro-fibrotic, and progesterone-resistant states, while fibroblast populations diversify into ECM-producing and angiogenesis-supporting subsets [ 80 – 84 ]. Fibroblast and ECM remodeling are especially important to lesion persistence. Fibrosis-rich and angiogenesis-associated fibroblast states have been identified across lesion types, and FAP-positive activated fibroblasts are detectable in endometriotic tissue, where they correlate with collagen deposition and infiltrating immune cells [ 83 , 84 ]. These data support a model in which lesion-associated fibroblasts act as local organizers of matrix stiffness, immune-cell recruitment, and vascular remodeling, thereby stabilizing a chronic wound-healing phenotype. At the soluble level, peritoneal fluid is enriched in inflammatory cytokines, chemokines, proteases, and extracellular vesicles that reinforce lesion-supportive signaling [ 76 , 77 , 85 , 86 ]. Endometriotic lesion-derived small extracellular vesicles can directly impair macrophage function through CD47/SIRPα-associated mechanisms, while additional exosome studies indicate that stromal-cell-derived vesicles promote M2-like macrophage polarization and reduced phagocytic capacity [ 85 , 107 – 109 ]. Annexin A2 and related peritoneal-fluid proteins have also emerged as candidate markers of microenvironmental remodeling, and annexin A2 suppression in peritoneal macrophages is mechanistically linked to impaired clearance of ectopic cells [ 86 , 110 ]. Together, these findings define the peritoneal cavity as an immunologically and stromally reprogrammed niche that actively supports implantation, persistence, and progression. Among lesion phenotypes, ovarian endometriomas are biologically distinctive because they develop within a semi-enclosed hemorrhagic cavity that accumulates heme, free iron, reactive oxygen species (ROS), inflammatory mediators, and degraded blood products over time [ 87 , 88 ]. This iron-rich microenvironment is increasingly viewed as a central driver of oxidative injury, metabolic stress, fibrosis, and reduced reproductive competence. Reviews focused on ovarian endometriomas emphasize that chronic low-grade inflammation, repeated cyclic bleeding, and oxidative stress alter the surrounding ovarian cortex and follicular compartment, even when gross disease appears localized [ 87 , 88 ]. Mechanistic work further shows that iron overload can disrupt granulosa-cell steroidogenic signaling through ROS/HIF-1α/FSHR/CYP19A1 circuitry, offering a direct pathway from biochemical injury to impaired follicular support and compromised oocyte quality [ 89 ]. This ovarian niche is not only inflammatory but also mechanically and transcriptionally specialized. Single-cell and spatial studies show that ovarian lesions are enriched for stromal, epithelial, and immune programs linked to oxidative stress adaptation, angiogenesis, and matrix remodeling [ 80 – 83 ]. These changes likely explain why ovarian endometriomas often have disproportionate effects on fertility relative to lesion size alone. The endometrioma environment may impair folliculogenesis through paracrine inflammatory spread beyond the cyst wall, alter local endocrine signaling, and accelerate tissue senescence or fibrosis in adjacent cortex [ 87 – 89 ]. In this sense, ovarian endometriomas represent a paradigmatic example of how lesion microenvironment can drive both local organ dysfunction and broader translational concerns, including disease recurrence, infertility, and, in selected cases, cancer-relevant progression biology. Deep infiltrating endometriosis (DIE) is characterized by dense fibrosis, smooth-muscle infiltration, vascular remodeling, and rich innervation, and is strongly associated with chronic pelvic pain and organ dysfunction. Compared with superficial peritoneal disease, DIE lesions appear to occupy a more mechanically constrained, nerve-rich, and inflammation-fibrosis–coupled niche [ 79 – 83 ]. Single-cell analyses have highlighted coordinated transcriptional programs linked to immunotolerance, angiogenesis, fibrosis, and stromal differentiation across eutopic and ectopic tissues, while newer atlases suggest that lesion-specific niche interactions sustain growth through crosstalk among stromal, mesothelial, immune, and endothelial compartments [ 79 – 82 ]. This biology is reinforced by translational data implicating IL-8 signaling in the coupling of inflammation and fibrosis, with anti-IL-8 blockade reducing lesion burden and fibrotic remodeling in preclinical models [ 91 ]. Pain in DIE is now best understood as a microenvironmental outcome rather than a simple measure of lesion bulk. Neuroendocrine-immune frameworks indicate that lesions develop reciprocal signaling loops among immune cells, inflammatory mediators, and nociceptive fibers, thereby promoting peripheral sensitization and persistent pain [ 92 ]. Histamine-associated inflammatory programs, mast-cell activation, and neurovascular remodeling further strengthen the link between tissue inflammation and symptom generation [ 97 , 98 ]. Collectively, peritoneal, ovarian, and deep lesions should be regarded as biologically distinct but overlapping microenvironments, each defined by a different balance of inflammation, fibrosis, vascularization, endocrine adaptation, and neural engagement.

Oral GnRH antagonists and combination regimens: Oral GnRH antagonists have now moved from late-stage development into routine clinical use in multiple settings and represent the most mature nextgeneration hormonal platform in endometriosis [ 145 – 150 ]. Across recent meta-analyses and reviews, these agents consistently reduce dysmenorrhea, non-menstrual pelvic pain, and, in some studies, dyspareunia, while offering practical advantages over GnRH agonists such as oral administration, rapid onset, avoidance of flare, and faster recovery of ovarian function after discontinuation [ 145 – 150 ]. Their main limitation remains the same feature that underlies their efficacy: dose-dependent estradiol suppression, which can produce vasomotor symptoms, lipid changes, and bone mineral density loss if left unbuffered [ 145 – 150 ]. The major pharmacologic innovation in this class has therefore been not simply GnRH antagonism, but controlled estrogen suppression. Add-back regimens-typically low-dose estradiol plus norethindrone acetate or related combinations-have emerged as the key strategy for widening the therapeutic window by reducing hypoestrogenic toxicity while preserving analgesic benefit [ 149 , 150 ]. This principle is now best exemplified by relugolix combination therapy, which integrates oral GnRH antagonism with built-in add-back in a single daily regimen. Pivotal and extension studies have shown clinically meaningful reductions in dysmenorrhea, non-menstrual pelvic pain, and dyspareunia, alongside improvements in functioning and quality of life and relative stabilization of bone health over longer follow-up [ 152 , 153 , 155 ]. These data are important because they shift the field from short-term suppression toward more realistic chronic-use frameworks, which is essential for a disease that often spans decades [ 149 , 152 , 153 , 155 ]. Linzagolix has reinforced a second important concept: dose flexibility. Phase 3 data support the efficacy of linzagolix versus placebo for endometriosis-associated pain, with higher-dose regimens plus add-back providing a practical means of balancing efficacy and safety [ 151 ]. This dose-titration logic aligns with a broader move toward more personalized endocrine suppression, in which the degree of estrogen lowering may eventually be adapted to symptom severity, tolerability, bone risk, and reproductive goals rather than applied uniformly [ 145 , 147 , 149 – 151 ]. In parallel, Japanese trial programs with relugolix continue to refine how this class compares with established progestin-based approaches such as dienogest and with depot GnRH agonists in real-world second-line use [ 154 , 156 ]. Overall, the field is moving from a simple “suppress ovarian function” paradigm toward a more nuanced hormonal engineering approach in which pharmacokinetics, reversibility, and long-term tolerability are as important as short-term analgesia. This hormonal progress is scientifically important beyond symptom control. Endometriosis is increasingly recognized as a lesion-contextual disease in which endocrine signals interact with inflammatory, fibrotic, and neurovascular circuits. By partially uncoupling efficacy from severe hypoestrogenism, newer antagonist-based strategies may provide a better platform for future combination therapy, including regimens that pair endocrine backbones with non-hormonal microenvironment-directed agents. That framework is especially relevant for ovarian and deep disease, where oxidative stress, fibrosis, and clonal lesion persistence may not be fully reversed by endocrine suppression alone. Selective progesterone receptor modulators (SPRMs) remain conceptually attractive because they aim to modulate progesterone signaling without necessarily inducing profound hypoestrogenism [ 157 – 160 ]. In principle, this makes them especially relevant to endometriosis, where progesterone resistance is central to lesion persistence and where a tissue-selective PR-directed strategy could, at least theoretically, combine antiproliferative, anti-inflammatory, and endometrium-modifying effects while preserving a more favorable systemic endocrine profile [ 159 , 160 ]. The challenge has been translating that pharmacologic elegance into clinically durable and safe therapies. Preclinical work on newer mesoprogestins continues to support this class as biologically viable. EC313, for example, has demonstrated anti-proliferative activity and favorable selective progesterone receptor modulation in fibroid models and preclinical profiling studies, suggesting that next-generation SPRM scaffolds may achieve stronger PR selectivity with fewer off-target liabilities than older agents [ 157 , 158 ]. Although these data are not endometriosis-specific clinical evidence, they support the broader idea that SPRM design has not reached its ceiling and that tissue-selective PR modulation remains a credible future direction for endometriosis drug development [ 157 , 158 ]. Clinically, vilaprisan offers the strongest recent proof of concept in endometriosis. In a terminated phase 2b randomized trial, vilaprisan showed signals of pain benefit over placebo with an acceptable short-term tolerability profile, sustaining interest in the class while also highlighting the unresolved questions that continue to constrain its development [ 159 ]. Those questions include long-term endometrial effects, hepatic safety concerns that historically limited broader SPRM enthusiasm, and the need to define which patient subsets are most likely to benefit. In parallel, epigenetic reviews of endometriosis continue to reinforce why SPRMs remain scientifically compelling: lesion biology is shaped not only by hormone levels, but also by stable chromatin and transcriptional programs that may alter PR responsiveness and downstream anti-inflammatory signaling [ 160 ]. In that sense, SPRMs and related mesoprogestins may still become important in endometriosis, but likely as part of a more biomarker-aware and patient-stratified strategy than earlier development programs envisioned. A major frontier in endometriosis therapeutics is the development of non-hormonal agents that reduce pain and lesion progression without inducing medical menopause or compromising fertility [ 161 – 170 ]. This shift is biologically rational because endometriosis lesions are sustained not only by ovarian steroids, but also by chronic inflammatory signaling, altered angiogenesis, stromal remodeling, oxidative stress, immune tolerance, and neuroimmune crosstalk. Many of these pathways are also recognizable from cancer and fibrotic-disease biology, which is one reason oncology-style target discovery and pathway-directed treatment logic are becoming increasingly relevant in endometriosis [ 160 , 161 ]. The translational challenge is to identify targets that are sufficiently central to lesion persistence to produce meaningful benefit, yet sufficiently safe for use in a largely young and often fertility-conscious population. Among non-hormonal concepts, cytokine- and chemokine-directed therapies remain among the most biologically persuasive. Reviews of non-hormonal treatment development consistently position inflammatory mediators as core drivers of lesion persistence and pain, especially when coupled to fibrosis and neurovascular remodeling [ 161 ]. A pilot randomized trial of anakinra demonstrated that IL-1 pathway inhibition is feasible in endometriosis and may improve symptom and inflammatory readouts, although the study also made clear that adequately powered, phenotype-aware trials are still needed before this approach can be positioned clinically [ 162 ]. Even more compelling from a lesion-biology standpoint is the IL-8 axis. In preclinical work, a long-acting anti-IL-8 antibody reduced inflammation and fibrosis in endometriosis, and this mechanism has now advanced into early human testing through the phase 1 AMY109 program [ 163 , 164 ]. This is one of the clearest examples of a translational ladder in endometriosis: target identified from lesion microenvironment biology, validated in high-value preclinical systems, and then advanced into early clinical development [ 163 , 164 ]. Beyond canonical inflammation, several non-hormonal directions converge on lesion-supportive biology that is highly relevant to chronicity and, in ovarian disease, mildly cancer-adjacent processes such as angiogenesis, matrix remodeling, oxidative adaptation, and clonal survival. Reviews in this space increasingly emphasize anti-fibrotic and anti-angiogenic strategies because these pathways are tightly linked to established lesions and organ dysfunction [ 161 ]. Another emerging axis is prolactin biology, which sits at the intersection of endocrine signaling, nociceptor sensitization, and female-predominant pain disorders. Prolactin has been implicated in endometriosis pain and in endometriosis–migraine comorbidity, while antibody-based prolactin targeting has shown proof-of-principle in female pain models [ 165 , 166 , 168 ]. These studies do not yet establish prolactin blockade as an endometriosis therapy, but they strengthen the broader concept that endocrine-immune-neural pathways may be tractable for symptom clusters extending beyond pelvic pain alone [ 165 , 166 , 168 ]. Drug repurposing is also gaining momentum as a practical route to shorten timelines and leverage known safety frameworks. Recent computational and experimental repositioning studies have identified novel targets such as ROR1 and have begun to test clinically accessible compounds in preclinical systems, illustrating how transcriptomics-guided prioritization can move candidate drugs toward disease-relevant validation [ 167 ]. This approach is especially attractive in endometriosis because it may allow the field to borrow from oncology, immunology, and pain medicine without requiring every program to begin from de novo chemistry. Finally, local and technology-enabled delivery platforms may become just as important as the active agents themselves. Nanotechnology and biomaterials reviews increasingly frame endometriosis as a disease in which site-selective drug delivery could improve lesion exposure while minimizing systemic toxicity [ 169 ]. Injectable or implantable hydrogels, nanoparticle systems, and hybrid depots may be especially useful for anti-inflammatory, anti-fibrotic, or endocrine payloads that would be difficult to sustain safely at systemic concentrations [ 169 , 170 ]. A notable example is the near-infrared-responsive adhesive hydrogel delivering letrozole and polydopamine, which achieved localized photothermal/endocrine effects in preclinical models [ 170 ]. Such systems are still early, but they reflect a broader trend: endometriosis drug development is increasingly moving from a “one drug, one endocrine mechanism” model toward a platform-based strategy integrating pharmacology, biomaterials, and lesion microenvironment biology. Taken together, the emerging pipeline suggests that the next phase of endometriosis therapeutics will not be defined by a single breakthrough class, but by layered strategies: optimized hormonal suppression where appropriate, selective PR modulation where biologically justified, and non-hormonal or delivery-enabled interventions targeting inflammation, fibrosis, angiogenesis, oxidative stress, and neuroimmune signaling. This evolution is scientifically important because it aligns treatment development more closely with the actual biology of lesions-including, in some contexts, biological features that overlap with chronic inflammatory and cancer-relevant microenvironmental programs. Biologics and platform-based therapeutics represent one of the most forward-looking areas in endometriosis drug development because they shift the field away from uniform endocrine suppression toward lesion- and pathway-directed intervention. This is particularly relevant for patients with refractory pain, fibrotic disease, or intolerance to hypoestrogenic states. It also introduces a mild but important cancer-relevant perspective: some of the same biological processes that sustain endometriotic lesions-immune evasion, cytokine dependency, stromal remodeling, angiogenesis, and oxidative adaptation-are also familiar from tumor-microenvironment biology. As a result, endometriosis is increasingly being approached with translational strategies that resemble those used in oncology, even though the disease itself remains histologically benign [ 171 – 177 ]. Among biologics, monoclonal antibodies and receptor-directed agents are the most clinically visible class. Recent overviews of antibody-related drug development in gynecology show that antibody platforms are expanding beyond oncology into chronic benign disorders, including endometriosis, as advances in Fc engineering, half-life extension, and target selectivity improve feasibility for long-term use [ 171 ]. In endometriosis, the best example is the IL-8 axis, which has emerged as a biologically coherent target because IL-8 links inflammation, angiogenesis, and fibrosis-three core features of persistent lesions. Early commentary on the first-in-human AMY109 program underscores both the promise and the caution of this approach: chemokine blockade may move the field toward lesion-modifying therapy, but infection risk, cost, patient selection, and the need for meaningful non-pain endpoints remain central translational challenges [ 172 ]. Similar logic applies to other biologic concepts aimed at endocrine-immune interfaces, including prolactin-related and receptor-targeted programs, which may ultimately prove useful in carefully selected endotypes rather than as universal therapies [ 171 , 172 ]. RNA-based therapeutics and gene-targeted approaches remain largely preclinical in endometriosis, but the enabling science around them is advancing rapidly. Recent reviews on microRNA technologies and miRNA drug delivery emphasize that oligonucleotide therapeutics are becoming increasingly realistic because of improvements in chemical stabilization, targeting, nanoparticle carriers, and stimuli-responsive delivery systems [ 173 , 174 ]. This is especially relevant to endometriosis, where epigenetic dysregulation, endocrine resistance, inflammatory persistence, and lesion-specific signaling programs could in principle be addressed through miRNA- or siRNA-based approaches. Here again, the cancer field is instructive: oncology has driven major advances in RNA therapeutics, and the lessons from miRNA drug development in cancer-particularly around off-target effects, immune activation, delivery barriers, and biomarker-guided patient selection-are directly informative for gynecologic translation [ 175 ]. Likewise, while CRISPR-based therapies remain conceptual for endometriosis, current reviews of genome editing in cancer and medicine show how rapidly the field is progressing in terms of delivery, specificity, durability, and regulatory thinking [ 176 , 177 ]. For endometriosis, the near-term value of CRISPR is more likely to be in target validation and mechanistic modeling than direct patient therapy, but it clearly expands the horizon of what may eventually be tractable. Novel delivery systems may ultimately be as important as the active agents themselves. Reviews of nanotechnology in reproductive healthcare emphasize the potential of nanoparticles, injectable depots, implantable systems, and tissue-targeted carriers to stabilize exposure, reduce pill burden, and concentrate drug at the lesion site while limiting systemic toxicity [ 178 ]. This principle is highly attractive in endometriosis because many promising agents-whether hormonal, anti-inflammatory, anti-fibrotic, or RNA-based-are limited by the consequences of chronic systemic exposure. Local or semi-local delivery to the peritoneal cavity, ovary, or uterus could therefore widen the therapeutic window and make combination strategies more realistic. In translational terms, endometriosis may benefit from the same platform logic increasingly used in oncology: pairing a biologically active agent with a delivery system that improves tissue selectivity and pharmacodynamic precision [ 175 – 178 ]. A complementary strategy is drug repurposing, which offers a practical bridge between mechanism discovery and near-term clinical development. Transcriptomics-based repositioning pipelines have already identified approved or clinically accessible compounds as candidate endometriosis therapies, supporting the idea that disease-signature reversal can generate plausible non-hormonal leads [ 179 ]. More recent work has strengthened this approach by identifying specific repurposing candidates, including simvastatin and primaquine, that attenuated pain-related behaviors and reversed endometriosis-associated transcriptional programs in preclinical models [ 180 ]. The value of this approach is not only speed and cost-efficiency; it also helps connect endometriosis to broader therapeutic ecosystems in oncology, immunology, and pain medicine, where many mechanistically relevant compounds have already cleared major safety hurdles. Going forward, the most compelling biologic and advanced-modality programs in endometriosis are likely to be those that combine strong lesion biology, feasible delivery, biomarker-linked patient selection, and patient-centered endpoints extending beyond pain alone.

For almost a century, Sampson’s retrograde menstruation theory has provided the dominant framework for endometriosis initiation: viable endometrial fragments reflux through the fallopian tubes during menstruation, implant on peritoneal surfaces, and establish ectopic lesions [ 16 – 19 ]. Contemporary evidence, however, suggests that lesion establishment is not a single implantation event but a dynamic, cycle-linked process requiring repeated episodes of ovulatory menstruation, recurrent tissue reflux, and a permissive host environment [ 16 , 17 ]. Anatomical determinants, including cervical outflow resistance, intramural tubal architecture, and myometrial abnormalities, may influence the volume and force of retrograde flow and thereby contribute to inter-individual variation in disease susceptibility [ 18 ]. These considerations help explain the long-recognized “retrograde menstruation paradox”: menstrual reflux is common, whereas clinically significant endometriosis develops only in a subset of exposed individuals [ 19 ]. Recent molecular studies have strengthened an implantation-and-selection model by showing that eutopic endometrium can contain clonally expanded epithelial populations with cancer-associated somatic alterations, including KRAS, PTEN, and ARID1A-related events, that may disseminate and contribute to lesion biology [ 12 , 21 – 22 ]. Clinical associations between obstructive Müllerian anomalies and markedly higher rates of endometriosis further support a causal role for increased retrograde exposure while also underscoring the importance of anatomy and tissue permissiveness [ 22 ]. Taken together, current data support retrograde menstruation as a major initiating mechanism for pelvic disease, but one that operates within a broader framework of clonal endometrial biology, host susceptibility, and microenvironmental selection. The stem/progenitor cell hypothesis refines classical implantation theory by proposing that a minority population of clonogenic endometrial cells, rather than differentiated shed cells, is particularly well suited to survive reflux, attach to ectopic surfaces, and initiate lesions [ 23 – 26 ]. Recent reviews emphasize the existence of epithelial and stromal stem/progenitor niches in the human endometrium and suggest that dysregulated recruitment, survival, and differentiation of these cells may contribute to endometriosis pathogenesis [ 23 – 25 ]. Additional support comes from studies identifying Sox2-positive stem-like cells in lymph nodes from women with endometriosis, consistent with the possibility of lymphovascular dissemination of progenitor-like populations [ 26 ]. Developmental timing hypotheses have extended this concept further. Neonatal uterine bleeding has been proposed as a potential early-life seeding mechanism for endometriosis, although recent studies remain mixed and suggest that this hypothesis is biologically plausible but not yet proven [ 27 , 28 ]. More broadly, current perspective articles argue that stem/progenitor dissemination should not be viewed in isolation, but rather as one component of a multi-mechanism model in which menstrual reflux, developmental biology, immune dysfunction, and tissue context all interact in subtype-specific ways [ 29 ]. Coelomic metaplasia posits that peritoneal mesothelium or subcoelomic mesenchyme can transform into endometrium-like tissue, providing explanatory power for endometriosis in settings where retrograde menstruation alone is insufficient, including some premenarchal, extra-pelvic, and congenital-anomaly-associated cases [ 30 – 33 ]. A recent systematic review and meta-analysis of Mayer–Rokitansky-Küster-Hauser syndrome variants showed that endometriosis is substantially more frequent in patients with functional endometrium in uterine remnants, supporting an exposure-dependent component while still leaving room for developmental mechanisms [ 30 ]. Thoracic and diaphragmatic endometriosis likewise continue to motivate multi-mechanism models, with recent reviews highlighting the probable interplay of implantation, fluid dynamics, dissemination, and local tissue remodeling in extra-pelvic disease [ 31 – 33 ]. Marked heterogeneity in risk, phenotype, and lesion behavior has driven intensive investigation into inherited and acquired susceptibility mechanisms. Large multi-ancestry genomic studies continue to expand the catalogue of reproducible endometriosis risk loci and support a polygenic architecture enriched for pathways related to hormone response, inflammation, and tissue remodeling [ 34 ]. Epigenetic dysregulation is increasingly recognized as a mechanism that can stabilize abnormal lesion phenotypes by linking inflammation, endocrine signaling, and altered gene expression states across eutopic and ectopic tissues [ 35 ]. At the hormonal level, progesterone resistance remains a central concept, with recent reviews emphasizing that impaired progesterone responsiveness is not merely a downstream consequence of inflammation but may be embedded within lesion biology through persistent receptor, chromatin, and transcriptional abnormalities [ 36 ]. The eutopic endometrium in women with endometriosis is increasingly regarded as “ground zero” for disease initiation: a biologically altered tissue state that predisposes shed cells to survive, adhere, and invade at ectopic sites [ 36 – 38 ]. In a mouse model of endometriosis progression, transcriptomic profiling demonstrated coordinated activation of estrogenic, inflammatory, angiogenic, and fibrotic pathways in ectopic lesions together with disease-stage–linked changes in the eutopic endometrium [ 37 ]. Human single-cell analyses have further shown that ectopic lesions exhibit distinct metabolic programs compared with paired eutopic endometrium, consistent with hypoxia adaptation, altered bioenergetics, and stress-tolerant lesion ecology [ 38 ]. These data support the view that disease initiation reflects not only the presence of refluxed tissue, but also intrinsic changes in the source endometrium that favor persistence and implantation. Mounting evidence implicates microbiome dysbiosis and mucosal barrier dysfunction as additional layers in endometriosis pathogenesis. Multi-compartment profiling studies have identified disease-associated microbial signatures across vaginal, oral, and stool compartments, supporting the concept that endometriosis may involve interconnected mucosal-immune disruption rather than a purely local pelvic process [ 39 , 45 ]. Experimental studies now suggest that gut microbiota–derived metabolites can directly shape lesion biology: acetate has been shown to ameliorate disease in model systems through JAK1/STAT3-mediated macrophage reprogramming, whereas butyrate can enhance ferroptosis sensitivity through FFAR2/PPAR-γ/PINK1/Parkin-mediated pathways [ 40 , 41 ]. In parallel, newer reviews position the gut microbiota as an upstream regulator of estrogen metabolism, immune tone, and inflammatory set points relevant to lesion initiation and persistence [ 42 , 43 ]. Mendelian-randomization studies further support the possibility of causal links between selected gut microbial taxa and endometriosis risk, although directionality and generalizability remain unresolved [ 44 ]. Collectively, these data support a multifactorial initiation model in which an intrinsically altered eutopic endometrium is shed into a genetically and epigenetically susceptible host within a microbiome-conditioned inflammatory environment that can shift the balance from clearance to implantation [ 39 – 45 ].

The contemporary endometriosis pipeline remains dominated by hormone-modulating strategies, but its center of gravity is gradually shifting. Earlier drug-development reviews described a field largely confined to empirical hormonal suppression and surgery, with few credible late-stage alternatives and limited integration of translational biology into program design [ 181 – 184 ]. More recent clinical-development analyses now show a broader portfolio in which oral GnRH antagonists anchor the late-stage landscape, while earlier-stage programs increasingly target inflammatory, neuroimmune, fibrotic, angiogenic, and delivery-related mechanisms [ 181 – 184 ]. In practical terms, the global pipeline is still “pyramidal”: a narrow top of approved or near-approved assets, a modest middle of phase II concepts, and a wider discovery base that remains vulnerable to attrition because of imperfect models, weak biomarkers, and heterogeneous patient phenotyping [ 182 – 184 ]. At the late-stage end, the most commercially and clinically mature assets are the oral GnRH antagonists, especially elagolix historically and, more recently, linzagolix and relugolix-based regimens. The phase 3 development of elagolix was a major inflection point because it demonstrated that oral, dose-dependent suppression of gonadotropin signaling could reproducibly improve dysmenorrhea and non-menstrual pelvic pain in large randomized studies [ 185 ]. Follow-on analyses further showed improvements in work productivity and health-related quality of life, helping move the field beyond pain scores alone and establishing the commercial logic for chronic oral therapy in endometriosis [ 186 , 187 ]. Linzagolix represents the next major iteration of this concept. Phase 2 and later development emphasized flexible estrogen suppression, with lower doses potentially suitable for chronic use and higher doses paired with add-back to preserve efficacy while limiting hypoestrogenic toxicity [ 188 – 190 ]. This “dose architecture” is highly relevant to industry because it provides more than symptom control; it offers a differentiated positioning strategy around tolerability, chronic-use potential, and patient segmentation [ 188 – 190 ]. Even within this hormone-dominant landscape, the underlying development philosophy is changing. Older endometriosis R&D was often framed as finding a more tolerable suppressive agent; newer programs increasingly resemble precision chronic-disease development, where efficacy must be balanced against bone health, vasomotor symptoms, mood effects, fertility goals, adherence, and real-world function [ 183 – 190 ]. This shift has mild relevance to cancer-style thinking. Endometriosis is not a malignancy, but the most interesting pipeline concepts now increasingly target lesion ecology-angiogenesis, inflammation, stromal remodeling, immune signaling, and oxidative adaptation-rather than ovarian hormone production alone. In that sense, the field is beginning to borrow from oncology’s emphasis on microenvironment-directed intervention, pathway stratification, and biomarker-informed development. The most dynamic part of the current pipeline is therefore not the approved hormonal segment, but the mid-stage and emerging non-hormonal space. Here, programs are being prioritized not because they suppress ovulation, but because they may alter lesion-supportive biology while remaining compatible with longer-term use and, potentially, fertility-sparing care. This includes cytokine and chemokine blockade, fibrosis-directed strategies, neuroimmune pain modifiers, and advanced delivery systems. Industry interest in these mechanisms has grown because they may ultimately support the kind of disease-modifying claims that hormonal agents have struggled to achieve [ 181 – 184 ]. At the same time, the bottlenecks are substantial: preclinical systems remain only partially predictive, patient populations are biologically heterogeneous, and regulatory endpoints still center more on pain reduction than on lesion regression, fibrosis prevention, or interception of long-term progression [ 183 , 184 ]. Overall, the industry pipeline in endometriosis is best understood as being in a transitional phase. The first generation of commercially successful assets has validated oral antagonism of the reproductive axis, but the long-term strategic value of the field likely depends on whether non-hormonal and combination approaches can deliver durable benefit with acceptable safety. The most competitive future programs will likely be those that combine strong mechanistic rationale, phenotype-aware development, and outcomes that matter in real life-function, recurrence, fertility preservation, and possibly, in selected ovarian contexts, attenuation of chronic inflammatory states that overlap with cancer-relevant lesion biology [ 181 – 190 ]. Early-stage and first-in-class programs illustrate where the endometriosis pipeline is becoming most mechanistically ambitious. In contrast to the late-stage portfolio, which remains dominated by endocrine suppression, preclinical and phase I assets increasingly target lesion-supportive biology itself, including inflammatory and chemokine signaling, stromal-fibrotic remodeling, oxidative stress, neuroimmune pain circuitry, and local drug-delivery platforms designed to improve tissue selectivity and reduce systemic burden [ 191 , 195 ]. This shift is scientifically important because it aligns drug discovery more closely with the complexity of lesion microenvironments and with the need for fit-for-purpose translational models and biomarker strategies in early development [ 192 , 194 , 195 ]. A mild cancer-relevant perspective is also emerging in this space: several first-in-class concepts are aimed at processes such as immune tolerance, fibrosis, angiogenesis, and oxidative adaptation that are not unique to endometriosis and are also recognized features of tumor-adjacent or premalignant tissue ecosystems, particularly in ovarian disease [ 191 , 193 ]. Accordingly, the table below highlights representative early-stage targets and platforms that may help move the field from suppressive symptom control toward mechanism-led, potentially disease-modifying intervention [ 191 – 195 ]. From a geographic and market standpoint, the next phase of endometriosis drug development will depend not only on new mechanisms, but also on whether trial populations and real-world datasets better reflect the true diversity of disease burden. Recent prevalence-focused work argues that endometriosis remains substantially under-ascertained across health systems, which has direct implications for market sizing, reimbursement planning, and site selection for future trials [ 196 ]. In parallel, large electronic health record analyses and validation studies show that endometriosis populations are clinically heterogeneous, with distinct comorbidity clusters and measurable differences between administratively identified and surgically confirmed disease, reinforcing the need for more rigorous phenotyping if late-stage evidence is to generalize across care settings [ 197 , 198 ]. Artificial-intelligence and EHR-based approaches are increasingly positioned as tools to improve classification, early detection, and computable phenotyping, thereby helping bridge the gap between academic discovery cohorts and scalable industry development programs [ 199 ]. Collectively, these findings suggest that pipeline maturation will benefit from broader geographic participation, harmonized endpoint definitions, and linked clinical-data infrastructure capable of supporting both external validity and payer-relevant evidence generation [ 196 – 199 ]. A second lesson from recent trial successes and failures is that future programs are more likely to succeed when they align with the actual biology and symptom architecture of endometriosis rather than with a narrow ovarian-suppression model. Digital-health studies now show that symptom tracking can capture day-to-day variability and heterogeneous disease trajectories that are largely invisible to episodic clinic visits, offering a more realistic framework for endpoint development in early- and mid-stage trials [ 200 , 201 ]. Randomized evidence further indicates that telehealth-delivered cognitive behavioural therapy can improve endometriosis-related quality of life and pain, underscoring the importance of integrating chronic-pain science and patient-centered functional outcomes into therapeutic development rather than relying exclusively on lesion or hormone metrics [ 202 ]. This is also where a mild cancer angle becomes useful: several higher-value endometriosis targets now converge on processes such as immune tolerance, stromal remodeling, oxidative adaptation, and persistent pain circuitry-features that are not unique to endometriosis and are also familiar from tumor-adjacent microenvironments. For industry, this supports a more modern development strategy in which lesion biology, symptom endotypes, and tissue context are considered together rather than treating all endometriosis as a uniform endocrine disorder. Strategically, the most promising future industry-academic collaborations will likely be those built around deep phenotyping, real-world evidence, digital tools, and patient co-design. Patient-priority studies consistently show that affected individuals want earlier diagnosis, better treatments, and more clinically relevant research agendas [ 203 , 204 ]. At the same time, new genetic and multi-ancestry omics studies continue to expand the biologic landscape of endometriosis, linking risk architecture to candidate pathways and creating a stronger basis for mechanism-led target selection and responder enrichment [ 205 ]. The implication is clear: the field now has an opportunity to build a more translational pipeline in which genomics, lesion biology, EHR-linked phenotypes, digital symptom trajectories, and patient-defined priorities are integrated from discovery through early clinical development. Such a framework would improve target validation, strengthen go/no-go decisions, and increase the likelihood that future therapies address outcomes that matter in real life-pain, function, recurrence, fertility preservation, and, in selected ovarian contexts, attenuation of chronic inflammatory states with cancer-relevant features [ 196 – 205 ].

Taken together, the evidence reviewed here suggests that endometriosis should increasingly be understood as a heterogeneous, chronic, systemic disorder in which endocrine, immune, stromal, neural, and metabolic programs interact to shape lesion behavior and clinical outcomes [ 206 – 214 ]. This broader framing helps explain why symptom burden, fertility effects, and treatment response can vary so widely across patients, and why empiric suppression alone is often insufficient for durable control [ 206 – 213 ]. It also supports a more ambitious therapeutic goal: not only reducing pain, but altering the biological trajectories that underlie lesion persistence, fibrosis, recurrence, and pain sensitization [ 208 – 214 ]. Somatic driver-like alterations alone are unlikely to define malignant potential because they are also detected in histologically benign lesions. Clinically meaningful progression probably requires convergence of clonal alterations with atypical histology, persistent oxidative injury, immune escape, metabolic adaptation, and a permissive ovarian microenvironment. Current evidence therefore supports risk-stratification research rather than routine molecular surveillance of all patients with endometriosis. The path forward will likely depend on integrating several layers of evidence that have too often been studied in parallel rather than together. Mechanistic studies, advanced models, non-invasive diagnostics, imaging, and clinical phenotyping are beginning to converge on a more precise view of disease heterogeneity [ 206 – 213 ]. In practical terms, this argues for a translational framework in which targets are prioritized from human tissue biology, tested in model systems matched to the specific question, and advanced using biomarker-informed early trials with clinically meaningful endpoints [ 208 – 214 ]. Such an approach may also help refine which patients are most likely to benefit from endocrine backbones, which may require multimodal or non-hormonal strategies, and which lesion contexts warrant closer molecular attention. In endometriosis, disease modification should be distinguished from symptomatic improvement and could be operationalized through sustained lesion regression or stabilization, reduced fibrosis and neuroinflammation, prevention of recurrence or repeat surgery, preservation of ovarian reserve and organ function, and durable benefit after treatment withdrawal. A mild cancer-related perspective may also be scientifically useful here. Endometriosis is not a malignancy, and most lesions will never progress to cancer. Nevertheless, selected lesion subtypes-particularly ovarian endometriomas and some deeply infiltrating lesions-can display persistent inflammatory remodeling, oxidative stress, fibrosis, angiogenesis, and clonal epithelial alterations that overlap with biological processes familiar from premalignant or tumor-adjacent tissue ecosystems [ 214 – 220 ]. Framed cautiously, this does not imply that endometriosis should be managed as a cancer precursor in routine practice. Rather, it suggests that concepts commonly applied in cancer research-microenvironment-aware target discovery, clonal lesion biology, biomarker-guided stratification, and disease-interception logic-may be informative for endometriosis as the field moves toward mechanism-based, fertility-sparing care [ 211 , 214 – 220 ]. The cancer association is strongest for ovarian endometriosis, whereas the relevance of deep infiltrating disease lies mainly in its locally invasive, fibrotic, neuroangiogenic, and apoptosis-resistant phenotype rather than in a clearly established risk of malignant transformation. Ultimately, future therapies will need to do more than temporarily suppress ovarian function. The most meaningful advances will likely come from multimodal strategies that are personalized to lesion phenotype, symptom mechanism, reproductive goals, and long-term health priorities. Achieving that goal will require sustained investment in women’s health research, broader and more representative clinical datasets, stronger industry-academic collaboration, and continued partnership with patients in defining outcomes that truly matter. If these elements can be aligned, the next decade may offer a realistic opportunity to move endometriosis closer to a model of integrated, biologically informed, and genuinely disease-modifying care. This narrative review integrates a rapidly evolving and methodologically heterogeneous literature. Many emerging mechanisms and therapeutic strategies remain supported primarily by cross-sectional human studies or preclinical models, and publication bias, inconsistent lesion phenotyping, and limited longitudinal validation should be considered when interpreting their translational readiness.

Three-dimensional scaffolds and biomaterials are becoming increasingly important in endometriosis research because they bridge the gap between oversimplified 2D culture and the structural complexity of human lesions. In contrast to flat plastic surfaces, biomaterial-based systems can reproduce key physical and biochemical features of the lesion niche, including extracellular matrix composition, stiffness, porosity, oxygen and nutrient diffusion, and the spatial organization of epithelial, stromal, endothelial, and immune cells. These properties are highly relevant to endometriosis, where lesion survival depends on attachment to mesothelial or ovarian surfaces, matrix remodeling, fibro-inflammatory signaling, and persistent hormone responsiveness. Recent bioengineering reviews emphasize that scaffold-based models are particularly useful when the biological question involves invasion, fibrosis, angiogenesis, drug penetration, or multicellular crosstalk rather than simple cell-intrinsic signaling [ 61 , 75 ]. Hydrogels have emerged as the most versatile scaffold class for modeling endometrium and endometriosis-like tissues because their hydration, mechanics, ligand presentation, and degradability can be tuned with relatively high precision. Natural hydrogels, including collagen, gelatin, hyaluronic acid, fibrin, and decellularized extracellular matrix derivatives, provide tissue-like biochemical cues and support glandular organization, stromal survival, and matrix remodeling. Synthetic and semi-synthetic systems, by contrast, offer higher reproducibility and better control over stiffness and signaling inputs, making them attractive for mechanistic studies and drug screening [ 62 – 67 ]. Defined hydrogel systems are especially valuable because conventional basement membrane extracts such as Matrigel suffer from batch variability and xenogeneic composition, both of which limit translational reproducibility [ 63 , 64 ]. In endometrial organoid work, fully synthetic extracellular matrices and hybrid hydrogels incorporating decellularized endometrial matrix have already shown improved support for epithelial-stromal crosstalk, glandular architecture, and physiologic hormone responsiveness [ 62 , 65 , 66 ]. These advances are directly relevant to endometriosis because they create a platform in which lesion epithelial cells, stromal fibroblasts, macrophages, or mesothelial cells can be co-cultured under controlled matrix conditions that better reflect human disease [ 61 – 67 ]. A major advantage of hydrogel-based modeling is the ability to recapitulate matrix-dependent lesion behavior. Endometriotic cells do not merely survive in ectopic locations; they interact dynamically with surrounding matrix to invade, recruit vessels, resist apoptosis, and promote fibrosis. Tunable hydrogels therefore allow investigators to ask how stiffness, adhesion motifs, protease sensitivity, and matrix composition alter lesion growth and cell-state transitions [ 61 , 64 , 67 ]. Endometrial and endometriotic organoids can be embedded in hydrogel systems that support epithelial polarity and secretory function, while stromal cells can be incorporated to model inflammatory and fibrotic remodeling [ 62 , 65 ]. More advanced platforms combine hydrogels with vascular and microfluidic elements, enabling the study of nutrient gradients, endocrine cycling, and paracrine signaling under dynamic flow [ 61 , 75 ]. Such approaches are especially useful for interrogating the early phases of lesion implantation and the later development of a chronic fibro-inflammatory niche [ 61 , 62 , 65 , 75 ]. Electrospun matrices represent a complementary biomaterial strategy with particular relevance to fibrosis-rich and mechanically remodeled lesion environments. Unlike bulk hydrogels, electrospun scaffolds are composed of micro- to nanoscale fibers whose diameter, alignment, porosity, and mechanical properties can be precisely tuned [ 68 , 70 , 71 ]. This fibrous architecture more closely resembles the collagen-rich extracellular matrix seen in chronic lesions and deep infiltrating disease. Electrospinning is also well suited to creating scaffolds with large surface area, high porosity, and cargo-loading capacity, making it attractive for both tissue modeling and local drug delivery [ 70 , 71 ]. Recent work in endometrial regenerative systems has shown that electrospun membranes can improve adhesion, survival, and bioactivity of human endometrial mesenchymal stromal/stem cells, while also modifying inflammatory and angiogenic gene programs [ 68 ]. Although most of this work has been performed in regeneration or repair settings rather than direct endometriosis models, the same design principles are highly relevant to modeling endometriotic stromal persistence, matrix invasion, and scaffold-enabled local therapy [ 68 , 70 , 71 ]. Another strength of electrospun systems is modularity. Fibers can be combined with hydrogels, biologic polymers, growth factors, or hormone-loaded nanocomponents to create hybrid matrices that mimic multiple aspects of the uterine or lesion environment at once [ 69 – 71 ]. For example, trilayer or composite scaffold systems have been developed for uterine tissue engineering, including constructs with controlled estradiol release and layered architectures designed to support epithelial, stromal, and vascular regeneration [ 69 ]. While these platforms were not developed specifically for endometriosis, they demonstrate how biomaterials can be engineered to deliver hormones, anti-inflammatory agents, or other payloads in a spatially controlled way. This is particularly relevant for endometriosis because local delivery remains a major translational goal: lesion-directed hydrogels, electrospun implants, or composite depots could theoretically improve local drug exposure while reducing systemic hypoestrogenic or off-target effects [ 61 , 69 , 71 , 75 ]. Importantly, biomaterial systems are beginning to move beyond generic “endometrial” models toward disease-specific matrices. Hybrid endometrial-derived hydrogels, decellularized endometrial extracellular matrix, and even lesion-derived scaffolds are now being explored as ways to preserve native biochemical signals that are lost in conventional culture [ 65 – 67 , 72 ]. In one particularly relevant example, human endometriosis-derived scaffold material improved cell seeding and engraftment in vitro and in vivo, suggesting that lesion-specific extracellular matrix is not just structural but biologically instructive [ 72 ]. Similarly, patient-derived 3D peritoneal endometriosis culture systems and spheroid-based mesothelial-clearance assays are helping connect biomaterial engineering to early lesion initiation and invasion, including phenomena that parallel ovarian cancer spheroid–mesothelium interactions [ 73 , 74 ]. These systems are valuable because they allow the investigator to separate “seed” from “soil” questions: whether lesion behavior is primarily driven by endometriotic cells themselves or by the matrix and mesothelial context into which they are introduced [ 72 – 74 ]. Despite these strengths, scaffold-based endometriosis models still have limitations. Many remain technically demanding, require substantial optimization, and are not yet standardized across laboratories [ 61 , 75 ]. Some systems excel at modeling epithelial architecture but underrepresent immune or neural compartments; others are mechanically sophisticated but biologically incomplete. Long-term hormonal cycling, lesion chronicity, and integrated pain biology remain difficult to reproduce. Even so, 3D scaffolds and biomaterials now provide one of the most promising routes toward next-generation endometriosis models, particularly when combined with organoids, primary stromal cells, immune co-culture, or microfluidic control [ 61 – 75 ]. Going forward, the most informative systems are likely to be hybrid platforms that unite tissue-matched matrix chemistry, tunable mechanics, multicellular organization, and tractable pharmacology, thereby enabling more faithful modeling of lesion initiation, fibrosis, and treatment response than either 2D culture or conventional transplantation models alone [ 61 , 62 , 65 , 72 – 75 ].

Endometriosis is a chronic, estrogen-dependent inflammatory disorder characterized by the presence of endometrial-like glands and stroma outside the uterine cavity and is increasingly recognized as a systemic, multi-organ disease rather than a purely pelvic condition [ 1 ]. Contemporary diagnostic guidance now emphasizes structured clinical evaluation, imaging where appropriate, and selective surgical confirmation, reflecting efforts to shorten time to diagnosis and reduce unnecessary delay [ 2 ]. Symptoms often begin in adolescence, with substantial effects on schooling, psychosocial development, and later reproductive health, yet delayed recognition remains common [ 3 ]. The disease also imposes substantial direct and indirect economic burden, including out-of-pocket costs for medications, fertility care, travel, and supportive management [ 4 ]. Beyond its clinical burden, endometriosis is now understood as a chronic inflammatory, fibrotic, and neuroimmune disorder in which lesion persistence is sustained by broad immune remodeling, altered host–microbe interactions, and lesion-specific stromal and metabolic adaptation [ 5 – 8 ]. Ovarian endometriosis is particularly relevant in this context because ovarian endometriomas display persistent inflammatory and remodeling programs, while iron overload and oxidative stress within the ovarian microenvironment may contribute to follicular dysfunction and infertility [ 7 , 8 ]. These long-term biological effects are mirrored by measurable life-course consequences, including adverse educational and occupational outcomes after early diagnosis [ 9 ]. Recent reviews also emphasize that current symptom-based care pathways continue to leave many patients with incomplete relief and fragmented support [ 10 , 11 ]. Although histologically benign, endometriosis exhibits several features that are highly relevant to cancer biology, including clonal epithelial expansion, somatic alteration-associated lesion evolution, invasive behavior, angiogenesis, fibrosis, and lesion-specific metabolic reprogramming [ 7 , 8 , 12 ]. These properties are especially important in ovarian endometriosis, where a chronically inflamed, iron-rich niche may provide permissive conditions for genomic stress and, in selected contexts, malignant transformation. At the same time, endometriosis remains underdiagnosed at the population level, with prevalence estimates varying substantially by ascertainment strategy and resource setting [ 13 ], and current treatment remains largely suppressive rather than disease-modifying, driving interest in non-hormonal and mechanism-based therapeutic strategies [ 14 ]. Against this backdrop of major unmet clinical need and rapidly evolving molecular insight, there is a strong rationale to reassess endometriosis not only as a gynecologic pain disorder but also as a clonal, microenvironment-driven disease with partial overlap with tumor-like biology. A central obstacle remains persistent diagnostic delay, which continues to span years across many health systems despite increasing awareness [ 15 ]. In this review, we therefore examine current concepts of disease initiation, lesion microenvironment, immune dysfunction, experimental modeling, and emerging therapeutics, while also highlighting how clonal lesion biology, ovarian inflammatory remodeling, and cancer-relevant molecular programs may help refine future strategies for precision treatment and interception of endometriosis-associated malignant progression.

Although endometriosis is histologically benign, it increasingly appears to occupy a biologically important interface between chronic inflammatory disease and cancer-relevant lesion evolution. Epidemiologic evidence supports a non-random association between endometriosis and selected malignancies, most consistently endometriosis-associated ovarian cancer (EAOC), particularly the clear cell and endometrioid histotypes, while the overall absolute risk of malignant transformation remains low [ 46 – 51 ]. Recent reviews and clinicopathologic syntheses emphasize that this association should not be interpreted as evidence that all endometriosis is premalignant; rather, it indicates that a subset of lesions, especially ovarian lesions, may acquire molecular and microenvironmental features that support progression along a benign-atypical-malignant continuum [ 50 – 53 ]. Atypical endometriosis is currently regarded as the most plausible morphologic precursor to EAOC. Contemporary systematic and scoping reviews converge on a model in which benign endometriosis, atypical endometriosis, and invasive carcinoma can coexist within the same ovary and may represent sequential stages of progression in a subset of patients [ 48 – 50 , 55 ]. This interpretation is reinforced by molecular-pathology studies showing shared alterations between atypical endometriosis and adjacent ovarian carcinomas, particularly in pathways governing chromatin remodeling, PI3K/AKT signaling, proliferation, and stress adaptation [ 49 , 54 , 55 , 59 , 60 ]. In practical terms, atypical endometriosis provides the strongest bridge between the “benign disease” framework and a cancer-biology framework, because it captures the point at which chronic inflammatory and fibrotic lesion biology begins to overlap with clonal selection and oncogenic pathway activation [ 48 , 49 , 55 ]. At the molecular level, the cancer angle in endometriosis is driven not by generic inflammation alone, but by the convergence of somatic alterations, clonal persistence, and a permissive lesion niche. Recent reviews highlight recurrent involvement of ARID1A, PIK3CA, KRAS, PTEN, and related signaling networks in the transition from endometriosis to EAOC, particularly in ovarian clear cell and endometrioid carcinomas [ 47 , 52 – 57 , 59 , 60 ]. These alterations are not sufficient on their own to define malignant transformation, because similar mutations can be detected in histologically benign endometriotic epithelium. However, their presence supports the concept that at least some endometriotic lesions are clonally evolving rather than static implants, and that malignant progression likely depends on a second layer of selection imposed by the ovarian microenvironment, immune escape, and persistent oxidative injury [ 47 , 52 – 57 ]. Notably, mTOR and PI3K/AKT pathway dysregulation, SWI/SNF-ARID1A disruption, and histotype-specific transcriptional reprogramming are increasingly viewed as major mechanistic links between endometriosis and EAOC [ 54 , 56 , 57 , 59 , 60 ]. The ovarian endometrioma is particularly important in this framework because it represents a confined, hemorrhagic, iron-rich niche in which repeated cyclic bleeding, free heme, reactive oxygen species, inflammatory mediators, and altered extracellular matrix can collectively promote genomic stress and adaptive survival programs. Several recent reviews specifically identify ovarian endometriosis as the lesion context most strongly associated with malignant transformation, and emphasize that clear cell carcinoma in particular can be understood as an inflammation-adapted, endometriosis-associated malignancy [ 50 – 56 ]. This interpretation is also supported by experimental work showing that combinations of ARID1A loss, KRAS activation, AKT pathway activation, and MYC-related programs can drive endometriosis-associated tumorigenesis in model systems [ 58 ]. Thus, the ovarian endometrioma should not simply be viewed as a severe benign phenotype; it is also the lesion subtype that most clearly links endometriosis to oncogenic signaling, metabolic rewiring, and malignant potential [ 50 , 53 , 56 , 58 ]. These observations have important translational implications. First, they support studying endometriosis not only as a pain and infertility disorder, but also as a model of clonal inflammatory disease in which somatic alterations, fibrosis, immune remodeling, and stromal-epithelial crosstalk shape lesion behavior. Second, they justify greater attention to biomarkers that may identify lesions at increased risk of progression, particularly within ovarian disease. Third, they suggest that therapeutic strategies borrowed from cancer biology-including pathway-directed targeting of PI3K/AKT/mTOR signaling, chromatin-remodeling vulnerabilities, inflammatory-fibrotic circuits, and lesion-specific metabolic adaptation-may be relevant not only to EAOC, but also to the broader goal of disease interception in endometriosis [ 47 , 52 – 60 ]. Importantly, this does not mean that endometriosis should be treated as a premalignant condition in routine clinical practice. Rather, the value of the cancer angle is that it provides a mechanistic framework for understanding why some lesions behave in a tumor-like manner and why ovarian endometriosis, in particular, deserves closer molecular and translational scrutiny [ 46 – 60 ].