# Evidence digest — 2026-10-23 (week 2026-10-16..10-23)

The day-bounded publication-date query
(`endometriosis[Title/Abstract]`, datetype=pdat, 2026-10-16..2026-10-23)
returned **23** records. An entrez-date supplement (datetype=edat,
2026-10-17..2026-10-23) returned 42 records, **22** of which were not in the
pdat set. The supplement starts the day after the 2026-10-16 digest's edat
window, which already ran through 2026-10-16. Of the 45 records screened,
**32** are endometriosis-relevant and appear below (1 carry-over, 12 dropped
as noise). No Cochrane update and no new guideline. The one new randomized
trial in the window is a surgical-adjunct study of vasopressin, and it did
not separate ovarian-reserve decline. What stands out is a **comorbidity
meta-analysis that rates its own evidence low or very low**, a **narrative
review that restates the limit of a normal scan**, a **retrospective
frozen-embryo-transfer association with a very large odds ratio**, and a
**supplement review whose own conclusion is that supplements should not
replace standard treatment**. Prior digest: 2026-10-16 (open as PR #51,
updated onto main after the 2026-10-19 data refresh). Every entry below comes
from a fetched PubMed record; PMIDs are live. Nothing here changes an
OpenEndo repurposing verdict, the Yselty or saliva-diagnostics knowledge
pages, or a patient-facing recommendation.

## Diagnostics and imaging

- **Non-invasive diagnosis, from imaging to biomarkers and AI** — narrative
  review, not a systematic review. The authors state that major guidelines
  now endorse a clinical and imaging-based diagnosis without mandatory
  laparoscopy. Expert transvaginal ultrasound and MRI are described as
  identifying ovarian endometrioma and deep endometriosis with high
  specificity and as mapping disease for surgery. A normal scan does not
  exclude endometriosis, because superficial peritoneal disease remains
  largely invisible. No blood, urinary, or endometrial biomarker is described
  as meeting criteria for clinical use. Circulating miRNA panels are
  described as having failed independent validation, and a salivary miRNA
  signature is described as still needing independent validation in a
  lower-prevalence setting. Radiomics and AI are described as investigational
  pending external validation, calibration, and demonstrated clinical
  utility. The abstract gives no new accuracy figures.
  [PMID 42885805](https://pubmed.ncbi.nlm.nih.gov/42885805/) — *why it
  matters:* a current restatement of the imaging-first limit already in
  digest 2026-10-09 (PMID 42833036) and of the miRNA-methods problem in
  digest 2026-10-16 (PMID 42854833). It is not a new guideline. The saliva
  Endotest note from digest 2026-09-18 (PMID 42749315) is unchanged: still
  not an established standard.
- **Structured versus free-text MRI reports for deep pelvic endometriosis** —
  single-centre randomized reader study. Twenty radiology residents were
  assigned to free-text or structured reporting; five became unavailable
  before report generation. The remaining 15 readers each reported the same
  22 pelvic MRI examinations (330 reports: 154 free-text, 176 structured).
  The reference was expert imaging consensus, not surgical histology.
  Structured reporting increased completeness by 8.1 percentage points (95%
  CI 4.1–12.1) and clinically weighted completeness by 5.6 percentage points
  (95% CI 1.3–10.0), and took 4.52 minutes longer (95% CI 3.05–5.99).
  Structured reports also carried more explicit negative discrepancies
  versus the expert reference (rate ratio 2.94, 95% CI 1.73–4.98), mainly
  explicit compartment-based negative statements. In an exploratory paired
  reading, two gynaecologic surgeons scored 44 reports and rated structured
  reports higher for perceived completeness, usefulness, and adequacy;
  clarity ratings were similar. [PMID 42891704](https://pubmed.ncbi.nlm.nih.gov/42891704/)
  — *why it matters:* a reporting-format comparison, not a test of whether
  MRI detects endometriosis. Higher completeness is not diagnostic accuracy.
  Single centre, resident readers, imaging reference standard.
- **Platelet-to-lymphocyte ratio and CA-125 for endometrioma** —
  retrospective single-centre case-control study of 645 people having
  laparoscopic surgery in 2020–2024: 372 with endometriosis and 273 controls
  with benign ovarian teratomas. Endometrioma was associated with higher
  platelet-to-lymphocyte ratio (PLR), higher CA-125, and higher products of
  CA-125 with the neutrophil-to-lymphocyte, monocyte-to-lymphocyte, and
  platelet-to-lymphocyte ratios. Stage IV had higher platelet counts and
  higher CA-125 products than stage III. Preoperative bilateral endometrioma
  and elevated CA-125 were independent factors for stage IV. The authors
  state that, for people with dysmenorrhoea or pelvic pain, CA-125 had
  better diagnostic value for ovarian endometrioma than PLR × CA-125. The
  abstract does not report an AUC. [PMID 42897934](https://pubmed.ncbi.nlm.nih.gov/42897934/)
  — *why it matters:* another internal blood-marker comparison whose control
  group is teratoma, not the broader differential of an adnexal mass. The
  authors' preference for CA-125 is not a new diagnostic standard. Low
  certainty for any staging claim.
- **Uterine corpus flexion angle and the question-mark sign** — mixed
  prospective-retrospective case-control ultrasound study. Uterine corpus
  flexion angle (UCFA) rose with revised ASRM stage and with pelvic-adhesion
  severity. In multivariable regression, adhesion score was the main
  associate of UCFA (β=0.441), while the uterine flexion angle was mainly
  associated with baseline uterine position (β=0.508). In an
  adenomyosis-free endometriosis subgroup of 166, adhesion score remained
  the main associate (β=0.457). A UCFA cutoff of ≥220° for the sonographic
  question-mark sign had AUC 0.958. Median UCFA was 179.0° in the benign
  control group and 178.0° in the normal control group. The abstract does
  not state the total sample size. [PMID 42902224](https://pubmed.ncbi.nlm.nih.gov/42902224/)
  — *why it matters:* a proposed numeric definition of a sign sonographers
  already describe subjectively. The cutoff and the AUC are internal. Not a
  reporting standard. Epub 2026-07-25; indexed this week.
- **Immune-coagulation nomogram for endometriosis-associated ovarian cancer
  versus endometrioma** — retrospective single-centre study of
  treatment-naive people having primary surgery, January 2015–February 2025.
  Histology used Sampson-Scott criteria: endometriosis-associated ovarian
  cancer n=49, ovarian endometrioma n=146. Median fibrinogen-to-albumin
  ratio was 0.10 versus 0.07, and median absolute monocyte count was 0.39
  versus 0.31×10⁹/L. A model using age, menopausal status, monocyte count,
  fibrinogen-to-albumin ratio, CA-125, and HE4 had an internal AUC of 0.932
  (95% CI 0.887–0.968), higher than HE4 alone (0.835), the
  fibrinogen-to-albumin ratio (0.786), CA-125 (0.769), and monocyte count
  (0.657). At a Youden threshold of 0.468, sensitivity was 75.5% and
  specificity 95.9% (PPV 86.0%, NPV 92.1%). Calibration was acceptable
  (Hosmer-Lemeshow p=0.225). The authors state that prospective multicentre
  validation is still required. [PMID 42902392](https://pubmed.ncbi.nlm.nih.gov/42902392/)
  — *why it matters:* the same class of result as the endometriosis-associated
  ovarian-malignancy model review in digest 2026-10-09 (PMID 42835309).
  Internal discrimination is not a usable test. It does not change care.
  Epub 2026-07-23; indexed this week.
- **SEED screening measure, items only** — participatory revision of
  ENDOPAIN-4D for Canadian young adults aged 18–25. Four Delphi phases.
  Sixty-five experts took part in the first round; fifteen took part in the
  working meeting and the second round. The result is a 22-item, five-domain
  questionnaire named Screening for Early Endometriosis Detection (SEED).
  No sensitivity, specificity, or cutoff is reported.
  [PMID 42898862](https://pubmed.ncbi.nlm.nih.gov/42898862/) — *why it
  matters:* this is the measure whose validation protocol was logged in
  digest 2026-10-16 (PMID 42869590). It is still not evidence that SEED
  detects endometriosis. Epub 2026-10-15.

## Surgery

- **Vasopressin during laparoscopic endometrioma excision** — randomized
  trial at one university centre, September 2018–July 2021. Seventy-five
  women with a unilateral ovarian endometrioma were stratified by baseline
  anti-Müllerian hormone and assigned to vasopressin-assisted cystectomy or
  control. The primary outcome was the change in anti-Müllerian hormone at
  12 months. Vasopressin shortened bleeding-control time (2.2±2.5 versus
  6.5±6.8 minutes, p=0.002). Cyst-stripping time was 8.9±7.9 versus
  14.4±13.7 minutes (p=0.054); the abstract calls that reduction
  significant, and the reported p value is not below 0.05. The 12-month
  decline in anti-Müllerian hormone did not differ (−0.96±1.11 versus
  −1.14±0.92 ng/mL, p=0.488). A mixed-effects model found no group-by-time
  interaction (p=0.816). Histology showed no significant difference in the
  thickness or proportion of excised ovarian tissue.
  [PMID 42890533](https://pubmed.ncbi.nlm.nih.gov/42890533/) — *why it
  matters:* a negative primary result. Shorter electrocoagulation time did
  not preserve the measured ovarian reserve. It is not a reason to adopt or
  abandon vasopressin as standard practice. Epub 2026-10-02.
- **Nerve-sparing surgery for deep infiltrating endometriosis** — systematic
  review and meta-analysis (PRISMA 2020; PubMed, Embase, Scopus, and LILACS,
  January 2000–August 2025). Sixteen studies, 1,623 women with surgically
  confirmed deep disease. Pooled conversion to open surgery 0.3%,
  intraoperative complications 1.5%, postoperative complications 9.8%,
  urinary voiding dysfunction 6.9%, low anterior resection syndrome 10.2%,
  and postoperative Female Sexual Function Index 22.3. Several functional
  outcomes were substantially heterogeneous. The authors state that the
  evidence does not support a conclusion of equivalence or superiority
  versus conventional surgery. [PMID 42896655](https://pubmed.ncbi.nlm.nih.gov/42896655/)
  — *why it matters:* a feasibility snapshot, not a comparative verdict.
  Nerve-sparing is not established here as the better operation.
- **Lower urinary tract symptoms in deep endometriosis** — narrative review
  (PubMed/MEDLINE, English, 2000–July 2026). In women with deep endometriosis
  who had preoperative urodynamics, pooled estimates were voiding
  dysfunction in 21% and detrusor overactivity in 15%, with substantial
  heterogeneity. The authors treat direct urinary-tract involvement, neural
  dysfunction, inflammation, and surgical injury as separate routes to
  symptoms, and they state that evidence for neuroinflammatory and
  nociplastic mechanisms is largely indirect or experimental. They write
  that nerve-sparing techniques should be considered when surgery is
  indicated and technically feasible.
  [PMID 42905159](https://pubmed.ncbi.nlm.nih.gov/42905159/) — *why it
  matters:* a mechanism map for urinary symptoms, with the authors' own
  surgical sentence attached. That sentence is their conclusion. It is not
  an OpenEndo recommendation, and PMID 42896655 does not show that
  nerve-sparing is superior.
- **Headache burden after endometriosis surgery** — single-centre
  observational study of 88 women who had endometriosis surgery between 3
  months and 5 years before enrolment. Participants completed the Headache
  Impact Test-6 and the 12-item Short Form by recalling their preoperative
  state and reporting their current state. Prospective preoperative
  questionnaires were not available. Headache-burden scores were lower, and
  physical and mental quality-of-life scores were higher, at the
  postoperative assessment than in the recalled preoperative period, in
  women with and without migraine (p<0.001). The differences were larger in
  the migraine group. [PMID 42907837](https://pubmed.ncbi.nlm.nih.gov/42907837/)
  — *why it matters:* a before/after comparison that depends on remembered
  preoperative scores. It is not evidence that surgery treats migraine.
  Epub 2026-07-27; indexed this week.
- **Catamenial pneumothorax, two cases** — case report plus literature
  review. The abstract frames recurrent perimenstrual pneumothorax as
  under-recognised and presents two cases. It does not give ages, recurrence
  counts, histology, or a treatment comparison.
  [PMID 42889145](https://pubmed.ncbi.nlm.nih.gov/42889145/) — *why it
  matters:* another thoracic-endometriosis case report after the cluster in
  digest 2026-10-16. Two patients are not a pathway. Epub 2026-10-01.
- **Unicornuate uterus, adenomyoma, renal agenesis, and deep endometriosis**
  — one case, a 39-year-old woman, gravida 1 para 1, previously recorded as
  uterus didelphys. MRI reclassified the uterus as unicornuate with a left
  cavitated non-communicating rudimentary horn (ESHRE/ESGE U4a) occupied by
  a 6 cm adenomyoma, contralateral adenomyosis, a deep endometriotic nodule
  to the left pelvic sidewall with rectal tenting and no transmural
  invasion, a left endometrioma, and an absent left kidney. Surgery and
  histology confirmed the horn, the adenomyoma, adenomyosis, and ovarian
  endometriosis. She recovered and the abstract describes her as
  asymptomatic. [PMID 42889881](https://pubmed.ncbi.nlm.nih.gov/42889881/) —
  *why it matters:* one imaging reclassification. It is not a new standard
  for Müllerian anomalies. Epub 2026-09-15.
- **Endometrioid carcinoma in an endometrioma, with a contralateral
  borderline tumour** — one case, a 43-year-old woman with endometriosis
  diagnosed about five years earlier. A left endometrioma grew from
  6.1×6.8×5.2 cm (April 2024) to 8.8×9×9.5 cm (April 2025); MRI was O-RADS
  5. CA-125 was 408.6 U/mL, CA 19-9 was 643.58 U/mL, and RMI-I was 1225.8.
  Histology showed grade 2 endometrioid adenocarcinoma arising in a left
  endometriotic cyst, FIGO IC1. The right ovary looked normal in the
  operating room and contained a borderline seromucinous tumour. Omentum,
  peritoneal cytology, and 31 lymph nodes were negative.
  [PMID 42903660](https://pubmed.ncbi.nlm.nih.gov/42903660/) — *why it
  matters:* a single oncology case. Interval growth and a suspicious scan
  are this patient's findings. They are not a screening rule for
  endometriomas. Epub 2026-10-07.
- **Postpartum hemothorax, thoracic endometriosis suggested** — one case, a
  34-year-old woman, day 8 after vaginal delivery. An initially clear
  effusion became a hemothorax. Thoracoscopy showed small hematoma-like
  lesions on the visceral pleura. Pleural-fluid CA-125 was markedly
  elevated and pigment-laden macrophages were present. The abstract calls
  this strongly suggestive of thoracic endometriosis syndrome and does not
  report endometrial glands and stroma on histology.
  [PMID 42907392](https://pubmed.ncbi.nlm.nih.gov/42907392/) — *why it
  matters:* a suggested, not histologically confirmed, thoracic case. One
  patient. Not a postpartum pathway.

## Mechanisms and biology

- **RNF34, PKM2, glycolysis, and a curcumin experiment** — RNF34 was lower
  in endometriotic tissue and inversely associated with pyruvate kinase M2
  (PKM2). RNF34 promoted PKM2 ubiquitination and reduced PKM2 stability,
  and it suppressed glycolysis, proliferation, migration, and invasion in
  stromal cells. In a mouse model, RNF34-deficient uterine tissue produced
  larger lesions, and PKM2 inhibition reduced that increase. Curcumin
  raised RNF34 and reduced lesion burden; depleting RNF34 weakened those
  curcumin effects. Preclinical (tissue, cells, mouse).
  [PMID 42892355](https://pubmed.ncbi.nlm.nih.gov/42892355/) — *why it
  matters:* a ubiquitin-glycolysis mechanism. The curcumin result is an
  experiment in that model, not a clinical supplement trial. It does not
  move any M3 repurposing verdict.
- **Peritoneal-fluid extracellular vesicles** — pilot case-control study.
  Small extracellular vesicles from peritoneal fluid of women with
  endometriosis (n=7), women without endometriosis having surgery for benign
  non-inflammatory conditions (n=7), and follicular fluid (n=3). Raman
  spectra differed between endometriosis and control peritoneal vesicles at
  891 cm⁻¹ (proteins) and 968 cm⁻¹ (lipids). Follicular-fluid vesicles were
  smaller and more rigid in this cohort. The authors call the findings
  hypothesis-generating and say no diagnostic application can be claimed.
  [PMID 42891160](https://pubmed.ncbi.nlm.nih.gov/42891160/) — *why it
  matters:* a methods pilot of 14 peritoneal samples. Not a test. Epub
  2026-07-22.
- **Rat autologous-transplant model, checked with elagolix** — video
  protocol. Endometrium sutured to the peritoneal wall of Sprague-Dawley
  rats formed cystic lesions in 18 of 20 animals at 21 days (90%). Elagolix
  at 100 mg/kg daily by oral gavage for 4 weeks reduced final lesion volume
  versus the model group (inhibition rate 86.7%) and lowered serum FSH, LH,
  and estradiol and peritoneal TNF-α, VEGF, and IL-6. Pain and behaviour
  were not measured. [PMID 42891579](https://pubmed.ncbi.nlm.nih.gov/42891579/)
  — *why it matters:* a methods paper. The 100 mg/kg figure is a rat dose in
  this protocol. It is not a human dose and it is not new evidence about
  elagolix in people.
- **Laparoscopic swine model** — video protocol for autologous endometrial
  transplantation by laparoscopy, with optional fluorescein nanoparticle
  labelling and later laparoscopic retrieval. No lesion-take rate and no
  treatment result are reported.
  [PMID 42891621](https://pubmed.ncbi.nlm.nih.gov/42891621/) — *why it
  matters:* a large-animal platform. Not a pathophysiology result. Issue
  date 2026-10-16 and absent from `weekly/2026-10-16.json`.
- **Acute stress and endometriosis pain** — narrative review (English
  abstract of a Hungarian article). The authors describe the stress-pain
  relationship as bidirectional and name cognitive appraisal, coping, and
  pain catastrophising. They suggest that acute psychological stress is
  more likely to affect short-term pain fluctuation through those
  psychological routes than through a direct effect, and they state that
  empirical evidence on the timing is limited. No effect size.
  [PMID 42893499](https://pubmed.ncbi.nlm.nih.gov/42893499/) — *why it
  matters:* a framework, not a trial of a psychological treatment.
- **Statins in female rodents, including induced endometriosis** —
  systematic review of preclinical rodent studies (PubMed, Scopus, Embase,
  Web of Science, LILACS, SciELO; SYRCLE risk of bias). In
  normocholesterolemic females, statin exposure was associated with ovarian
  morphological change, including impaired follicular dynamics. In rodents
  induced to have endometriosis, the abstract reports a reduction in the
  volume of endometriotic implants. It does not name the statin, the number
  of endometriosis studies, or a pooled effect. No hyperlipidemic rodent
  studies were found. [PMID 42900000](https://pubmed.ncbi.nlm.nih.gov/42900000/)
  — *why it matters:* a preclinical sentence on the statin thread that
  digest 2026-09-18 logged in humans (PMID 42668641). An unnamed implant
  reduction without an estimate does not change that observational status
  and is not an M3 verdict.
- **Gut microbiome, symptoms, and taste-receptor variants** — case-only
  study of 50 women with surgically confirmed endometriosis. Rectal-swab 16S
  profiles fell into three community types dominated by Bacteroides,
  Prevotella, or Escherichia-Shigella. Microbiome composition did not
  differ significantly by stage. Beta diversity was associated with
  dysmenorrhoea (R²=0.13) and stage (R²=0.12). Variants in TAS2R14 were
  associated with diversity measures. Lower abundance of short-chain-fatty-acid-producing
  Lachnospiraceae and Ruminococcaceae was associated with dysmenorrhoea.
  There is no control group without endometriosis.
  [PMID 42898479](https://pubmed.ncbi.nlm.nih.gov/42898479/) — *why it
  matters:* symptom correlations inside one case series. Not a diagnostic
  microbiome test and not a dietary prescription. Epub 2026-10-05.

## Epidemiology and burden

- **Endometriosis and systemic comorbidities** — systematic review and
  meta-analysis of observational studies (Web of Science and Medline through
  31 December 2025; CRD420251258716). Fifty-five studies met inclusion; 48
  were pooled. Thirty-two studies covered autoimmune diseases, with study counts that
  overlap across diseases (systemic lupus erythematosus 14, Sjögren's
  syndrome 8, rheumatoid arthritis 10, multiple sclerosis 5, celiac disease
  5, inflammatory bowel disease 4), plus 17 allergy studies, 13
  pain-syndrome studies, and 6 endocrine studies. In
  cohort studies the pooled risk ratios were 1.81 for systemic lupus
  erythematosus (95% CI 1.50–2.18), 1.54 for Sjögren's syndrome (1.40–1.71),
  and 1.71 for rheumatoid arthritis (1.40–2.09). In case-control and
  cross-sectional studies the pooled odds ratios were 3.93 (2.59–5.98),
  3.13 (1.10–8.90), 2.27 (1.75–2.95), 2.30 (1.57–3.37), and 4.50 (1.94–10.42)
  for those first five, respectively. The inflammatory-bowel-disease odds
  ratio was 2.82 (95% CI 0.58–13.61); that interval includes 1, so it is not
  a demonstrated increase even though the abstract lists it with the
  significant odds ratios. Allergy and pain associations that excluded 1
  included asthma (RR 1.27, 1.14–1.40), eczema (OR 1.58, 1.34–1.87),
  fibromyalgia (RR 1.65, 1.42–1.92), chronic fatigue syndrome (RR 1.41,
  1.31–1.51), and irritable bowel syndrome (RR 2.51, 1.55–4.07; OR 3.49,
  3.21–3.80). Hypothyroidism (OR 1.89, 0.61–5.90) and hyperthyroidism (OR
  0.95, 0.60–1.52) were not significant. GRADE quality was generally low or
  very low because of bias in the included designs. The authors argue for
  treating endometriosis as a systemic immuno-inflammatory disease and for
  multidisciplinary care. [PMID 42901254](https://pubmed.ncbi.nlm.nih.gov/42901254/)
  — *why it matters:* the largest association map this week, and the authors
  grade it as weak. Associations are not mechanisms, and the practice
  sentence is theirs. This digest does not adopt a new disease definition
  or a new care model.
- **Obstetric and neonatal outcomes** — retrospective single-centre cohort
  of singleton deliveries, 1991–2021, excluding chromosomal abnormalities
  and congenital malformations. Of 232,476 deliveries, 224 (0.1%) had
  documented endometriosis. After generalized-estimating-equation adjustment
  for maternal age, gravidity, and fertility treatment, endometriosis
  remained associated with caesarean delivery (adjusted OR 2.14, 95% CI
  1.62–2.83). Other obstetric and perinatal outcomes that differed before
  adjustment were not independently associated. Perinatal mortality was
  comparable in the unadjusted comparison.
  [PMID 42894015](https://pubmed.ncbi.nlm.nih.gov/42894015/) — *why it
  matters:* a caesarean association in a cohort whose recorded endometriosis
  prevalence (0.1%) is far below the usual reproductive-age figure, which
  is a warning about under-ascertainment. One centre. Not a delivery
  protocol. Epub 2026-07-22.
- **Information needs in Australia** — cross-sectional anonymous online
  survey, convenience sample via Facebook and women's-health newsletters.
  Women aged 18 or older, living in Australia, with diagnosed or suspected
  endometriosis. Of 265 analysed participants from seven states, the highest
  "strongly agree" information needs were comorbidities (76.98%, 204/265),
  endometriosis clinics (75.85%, 179/236), and how to deal with dismissal
  by health professionals (75.54%, 176/233). Open answers also asked for
  pathogenesis, management including alternatives, public-versus-private
  care, and costs. [PMID 42893505](https://pubmed.ncbi.nlm.nih.gov/42893505/)
  — *why it matters:* a map of what this self-selected group says it cannot
  find. Denominators change by item. It is not an evaluation of an
  information resource. Epub 2026-10-18.
- **Food and supplement spending in Hungary** — cross-sectional mobile-app
  survey, December 2023–August 2024, 90% response. 1,645 women aged 14–45
  years (524 with endometriosis, 1,121 without). Women with endometriosis
  had smaller households (mean 2.49 versus 2.95) and were more often in the
  higher spending bands: over €400/month on food in 30% versus 24% (median
  band €301–€400 versus €251–€300), and supplement costs over €35/month in
  22% versus 9% (median band €16–€25 versus €6–€15). Diet adherence and
  supplement use were also higher. Endometriosis status remained associated
  with food and supplement expenditure after sociodemographic adjustment.
  [PMID 42898086](https://pubmed.ncbi.nlm.nih.gov/42898086/) — *why it
  matters:* a spending difference. It is not evidence that a special diet or
  a supplement changes symptoms. Epub 2026-09-28.
- **What endometriosis randomized trials have been powered to see,
  1990–2024** — bibliometric review of 536 endometriosis randomized trials
  with an English abstract, identified in Dimensions.AI. Medical-management
  trials were 55.6% (n=298), of which 177 involved a GnRH analogue.
  Complementary-therapy trials rose from 1 of the medical-management trials
  in 1990–1999 (0.847%) to 31 (28.7% of medical-management trials) in
  2020–2024. The United States and China contributed 70 and 65 trials.
  Among 445 trials with at least one continuous outcome, 55.7% (n=248) were,
  at 90% power, sized only for a standardised mean difference the authors
  class as large or greater. [PMID 42907689](https://pubmed.ncbi.nlm.nih.gov/42907689/)
  — *why it matters:* a description of the trial literature, including a
  power limit. It is not itself a treatment result. Epub 2026-07-27.

## Pharmacology and supportive care

- **Estetrol 15 mg with drospirenone 3 mg** — systematic review (Cochrane
  methods, PRISMA; PubMed, Embase, CENTRAL; no language or date limit). Two
  randomized trials and two observational cohorts, 198 patients in total.
  Quantitative pooling was not done. Versus placebo, the combination reduced
  pelvic pain; versus other active hormonal regimens, no significant
  difference was observed. Dysmenorrhoea improved on the authors' reading.
  Quality of life, sleep, and global ratings improved. Dyspareunia results
  were inconsistent. The most common adverse effect was intermenstrual
  bleeding or spotting. The authors say superiority over other combined
  hormonal contraceptives remains uncertain.
  [PMID 42892004](https://pubmed.ncbi.nlm.nih.gov/42892004/) — *why it
  matters:* a small, unpooled set. "Promising" is the authors' word. It is
  not a new first-line recommendation and it does not displace a GnRH
  antagonist or a progestin. Epub 2026-10-15.
- **GnRH-agonist plus aromatase-inhibitor pretreatment before frozen embryo
  transfer** — multicentre retrospective cohort, 2018–2024, at Northwestern
  University, the University Fertility Clinic in Skive, Denmark, and IVI
  Rome. Eligible records had endometriosis, adenomyosis, or both, and at
  least one previous failed conventional frozen transfer. Fifty-one patients
  contributed 217 cycles: 159 conventional and 58 pretreatment cycles
  (about two months of GnRH agonist with or followed by an aromatase
  inhibitor). Most conventional cycles were programmed hormone-replacement
  cycles (119/159, 75%). Pretreatment was associated with higher odds of
  live birth (OR 26.7, 95% CI 9.48–75.0). The authors say prospective
  studies are needed. [PMID 42892070](https://pubmed.ncbi.nlm.nih.gov/42892070/)
  — *why it matters:* a Danish site is in the cohort, and the odds ratio is
  the kind of figure that gets repeated without the design. The people in
  the pretreatment cycles had already failed conventional transfer, the
  analysis is retrospective, and the interval runs from about 9-fold to
  75-fold. That is not a protocol to adopt.
- **Nutritional supplements for endometriosis symptoms** — systematic review
  of 24 clinical trials, 2,019 women. Some trials reported improvements in
  pain and quality of life, particularly with formulated curcumin or with
  selenium, alone or with dienogest. Some fertility-related measures
  improved; pregnancy rates did not. Some combination treatments reduced
  ovarian endometriosis; supplements did not reduce recurrence after
  surgery. The authors' conclusion is that supplements may support standard
  treatment and should not replace it. The PubMed abstract does not report
  pooled effect sizes. [PMID 42900597](https://pubmed.ncbi.nlm.nih.gov/42900597/)
  — *why it matters:* the patient-facing item this week. The review's own
  limit is the one to keep: this is not a reason to replace hormonal or
  surgical care with a supplement, and it is not an OpenEndo recommendation
  to start curcumin, selenium, or any other product. The curcumin experiment
  in PMID 42892355 is a mouse finding and is a different record.
- **Dopamine D2 agonists as a non-hormonal idea** — narrative review of
  cabergoline and quinagolide. The proposed mechanism is reduced VEGF-A
  signalling and less aberrant angiogenesis. A randomized, double-blind,
  placebo-controlled pilot of cabergoline for 6 months reduced chronic
  pelvic pain scores without serious adverse events; the abstract gives no
  sample size. A post hoc look at quinagolide found regression in 61.3% of
  lesions, with ovulatory function preserved in 93% of participants. The
  authors say adequately powered phase 3 trials are still needed.
  [PMID 42903633](https://pubmed.ncbi.nlm.nih.gov/42903633/) — *why it
  matters:* early clinical signals without the numbers a decision would
  need. Cabergoline and quinagolide are not OpenEndo M3 candidates, and
  this review does not add them. Epub 2026-07-22.
- **Diet as support, narrative review** — no new cohort. The abstract
  discusses an anti-inflammatory pattern (omega-3 fatty acids, cruciferous
  vegetables, fibre) as distinct from routine food elimination, plus
  selected supplements and probiotics, and states that diet should be
  treated as supportive rather than as causal treatment. It gives no study
  count and no effect size. [PMID 42886549](https://pubmed.ncbi.nlm.nih.gov/42886549/)
  — *why it matters:* a narrative companion to the supplement review and to
  the low-FODMAP meta-analysis in digest 2026-10-16 (PMID 42876029), which
  pooled post-intervention scores rather than controlled effects. Not a
  diet to prescribe. Issue date 2026-10-16 and absent from
  `weekly/2026-10-16.json`.

## Notable clinical signals

- **No new guideline, Cochrane update, or regulatory decision** this week.
  Nothing here changes an OpenEndo repurposing verdict. Nothing here is
  adopted as a patient-facing treatment recommendation. Knowledge pages for
  Yselty and for saliva diagnostics are unchanged.
- **Comorbidity meta-analysis (PMID 42901254)** is the broadest association
  paper: higher pooled estimates for several autoimmune, allergic, and pain
  conditions, an inflammatory-bowel-disease interval that includes no
  effect, thyroid associations that are not significant, and a GRADE rating
  of low or very low. The authors' call for a systemic disease model and
  multidisciplinary care is their interpretation. It is not a new care
  pathway.
- **Non-invasive diagnosis (PMID 42885805)** repeats a limit that matters
  for reading a scan: a normal expert ultrasound or MRI does not exclude
  superficial disease, and no circulating or endometrial biomarker is
  described as ready for clinical use, including salivary miRNA. That is a
  bound on what current tests can claim. It is not a new pathway.
- **Frozen-embryo-transfer pretreatment (PMID 42892070)** reports OR 26.7
  for live birth after a GnRH-agonist plus aromatase-inhibitor lead-in, in
  a retrospective series of 51 patients and 217 cycles at three centres,
  one of them in Skive. The design (already-failed conventional transfers)
  and the width of the interval are the result. It does not change
  embryo-transfer practice.
- **Supplements (PMID 42900597)** and the diet narrative (PMID 42886549)
  are the records most likely to be read as advice. The supplement review
  covers 24 trials and 2,019 women and states that supplements should not
  replace standard treatment. The diet paper gives no effect size. Neither
  record is a recommendation to start a product or a diet.
- **Estetrol/drospirenone (PMID 42892004)** is 198 patients across four
  studies, unpooled, with no significant difference versus other active
  hormonal regimens. **Dopamine agonists (PMID 42903633)** are a pilot and
  a post hoc analysis without a phase 3 trial. Neither is a new option to
  adopt.
- **Vasopressin (PMID 42890533)** is the week's randomized trial: less time
  controlling bleeding, no difference in 12-month anti-Müllerian hormone.
  **Nerve-sparing (PMID 42896655)** reports complication and function rates
  without a superiority claim; the urinary-symptom review (PMID 42905159)
  still says nerve-sparing should be considered when feasible. That remains
  the authors' sentence.
- **Two model papers are internal only:** the question-mark angle (PMID
  42902224, AUC 0.958, total n not stated) and the ovarian-cancer nomogram
  (PMID 42902392, internal AUC 0.932, authors: multicentre validation still
  required). **SEED (PMID 42898862)** is a 22-item draft, not a validated
  screen. **CA-125 versus a platelet ratio (PMID 42897934)** uses teratoma
  controls and reports no AUC.
- **Case reports** (PMIDs 42889145, 42889881, 42903660, 42907392) are one
  or two patients. The postpartum hemothorax is suggestive, not
  histologically confirmed in the abstract. The carcinoma case is one
  person's imaging and markers. None of these is a pathway.
- **RNF34/curcumin (PMID 42892355)**, the rat elagolix protocol (PMID
  42891579), the swine protocol (PMID 42891621), peritoneal vesicles (PMID
  42891160), the microbiome series (PMID 42898479), and the statin-rodent
  review (PMID 42900000) are preclinical or case-only. The human statin
  observation from September is not upgraded.

## Erratum with a known text

- **Menstrual-fluid extracellular-vesicle RNA-seq table** — PubMed records
  this as an erratum for PMID 41924633 (Hum Reprod Open, DOI
  10.1093/hropen/hoag020). The PubMed abstract says only that it corrects
  that article. The Europe PMC full text of the correction (PMC13679381)
  states that the RNA-seq data deposited as NCBI GEO GSE310627, and shown in
  the original Supplementary Table S1, came from menstrual-fluid mesenchymal
  stem cells rather than from menstrual-fluid small extracellular vesicles,
  because two files from the same cohort had similar names. GEO and the
  table have been replaced with the vesicle dataset. The correction states
  that this does not change the results, analyses, conclusions, or
  scientific message of the original paper. The original paper was not
  re-read; it is outside this window and not in the weekly snapshots.
  [PMID 42899171](https://pubmed.ncbi.nlm.nih.gov/42899171/) — *why it
  matters:* a data-file mix-up that the publisher says leaves the paper's
  claims in place. Those claims were not re-checked here.

## Carried over / follow-ups checked

- **Carry-over, not re-counted:** PMID 42866613 (endometriosis and chronic
  pain, stress mechanisms) is in `weekly/2026-10-16.json` and was
  synthesized in digest-2026-10-16. The pdat query returned it again. The
  record was not re-analyzed.
- **No other PMID in this included set** appears in `weekly/2026-09-02.json`,
  `weekly/2026-09-04.json`, `weekly/2026-09-11.json`,
  `weekly/2026-09-18.json`, `weekly/2026-09-25.json`,
  `weekly/2026-10-02.json`, `weekly/2026-10-09.json`, or
  `weekly/2026-10-16.json` — checked programmatically.
- **Issue date 2026-10-16 overlaps the previous window's end date.** PMIDs
  42885805, 42886549, 42891579, and 42891621 are included above because they
  are absent from `weekly/2026-10-16.json`. PMID 42866613, also dated
  2026-10-16, is the carry-over.
- **Indexed this week with earlier or non-day publication dates, included
  above because they were not in prior snapshots:** PMIDs 42889145 (epub
  2026-10-01), 42889881 (2026 Sep), 42890533 (epub 2026-10-02), 42891160
  (epub 2026-07-22), 42892004 (2026-10-15), 42892070 (2026-09-21), 42893505
  (publication date "2026 Jan-Dec"), 42894015 (epub 2026-07-22), 42898086
  (2026 Nov; epub 2026-09-28), 42898479 (epub 2026-10-05), 42898862 (epub
  2026-10-15), 42902224 (epub 2026-07-25), 42902392 (epub 2026-07-23),
  42903633 ("2026 Oct", epub 2026-07-22), 42903660 (2026 Dec; epub
  2026-10-07), 42907689 and 42907837 (epub 2026-07-27).
- **Dropped as noise (12):**
  - PMID 42889946 — cost-to-baby and national ART utilisation across 22
    countries. Endometriosis appears because a coauthor is identified as
    President of the World Endometriosis Research Foundation, in the
    disclosures. The analysis is not an endometriosis analysis.
  - PMID 42891460 — pelvic MRI in 262 people without sexual experience
    (63 adolescents). Endometriosis is an author keyword. The abstract
    reports organic versus functional findings and does not give an
    endometriosis count.
  - PMID 42892851 — cumulative MUSA features and adenomyosis severity in
    Kinshasa, 211 women. Endometriosis is an adjustment covariate
    ("recorded concomitant endometriosis"), not the disease under study.
  - PMID 42895300 — organoid and assembloid review of female reproductive
    biology. Endometriosis is one of three modeled conditions. The abstract
    gives no endometriosis-specific model count or result.
  - PMID 42895771 — absorption, distribution, metabolism, and excretion of
    linzagolix in rats and monkeys. Endometriosis is named as an indication
    alongside uterine fibroids. There is no endometriosis experiment. The
    Yselty knowledge page is unchanged.
  - PMID 42896814 — bisphenol A and hormone-associated disorders, 44
    studies. Endometriosis is a keyword. The pooled results in the abstract
    are polycystic ovary syndrome, breast cancer, and benign endometrial
    disorders. No endometriosis effect size is reported.
  - PMID 42897337 — prophylactic cefazolin and surgical-site infection
    after laparoscopic adnexal surgery, 1,678 patients. Endometriosis is a
    propensity-score covariate. Infection was 1.1% with and without
    cefazolin.
  - PMID 42897519 — serum androstenedione and dysmenorrhoea, 139
    premenopausal participants. The association the abstract reports is in
    women without endometriosis. No endometriosis-subgroup effect size is
    given.
  - PMID 42900601 — oxidative stress, infertility, and medicinal plants.
    Endometriosis is one listed cause of female infertility. The abstract
    gives no endometriosis-specific result.
  - PMID 42900624 — reversible camptocormia attributed to deep infiltrating
    endometriosis. The PubMed record has no abstract, so the outcome is not
    verifiable from the record.
  - PMID 42904152 — narrative review of paediatric and adolescent
    gynaecologic ultrasound. Endometriosis is named as one condition for
    which adult-derived criteria are still used. No endometriosis cohort or
    accuracy figure.
  - PMID 42905530 — letter to the editor on a nerve-and-muscle stimulation
    paper. The PubMed record has no abstract.

---

*Sources: PubMed snapshot `weekly/2026-10-23.json` (E-utilities esearch +
esummary + efetch). Primary: `endometriosis[Title/Abstract]`, datetype=pdat,
2026-10-16..2026-10-23 (23 records). Supplement: datetype=edat,
2026-10-17..2026-10-23 (42 records, 22 not in the pdat set). 45 records
screened, 32 included, 1 carry-over, 12 dropped. The menstrual-fluid
erratum text was read from the Europe PMC full text of PMC13679381; the
PubMed record itself only names the corrected DOI. Synthesized by Galahad,
2026-10-23. Research infrastructure, not medical advice.*