EGF as a non-invasive biomarker of endometriosis: a case control study Authors: Emma R Rashes Gertel, Cassandra C Daisy, Katherine Kaplan, Steven J Staffa, David Zurakowski, Allison F Vitonis, Kathryn L Terry, Amy L Shafrir, Simon T Dillon, Stacey A Missmer, Michael S Rogers, Towia A Libermann, Marsha A Moses Published: 2026 ABSTRACT AbstractEndometriosis is a chronic inflammatory condition that affects an estimated 1 in 10 women but is often mis- and underdiagnosed due to its non-specific symptoms and the lack of a non-invasive diagnostic test. This study aimed to identify and validate a potential non-invasive biomarker for endometriosis. This study applied quantitative proteomics discovery approaches to identify and validate non-invasive biomarkers of endometriosis with a long-term goal of leveraging them for purposes of diagnosis and therapeutic monitoring. Isobaric tags for relative and absolute quantification (iTRAQ) combined with mass spectrometry were used to identify and quantitate proteins and peptides in urine samples from participants with surgically-confirmed endometriosis (n = 73) and 1:1 age-matched participants never diagnosed with endometriosis (n = 73). Among those aged 25–47 at urine collection, Epidermal Growth Factor (EGF) (validated using monospecific enzyme-linked immunosorbent assays (ELISA)) was present at significantly lower levels in the urine of participants with surgically-confirmed endometriosis compared to controls (P = 0.02) and had an excellent negative predictive value (NPV) of 94.3% with a cutoff of > 800 pg/ug as determined by Bayes’ formula. Urinary EGF levels were also significantly lower in the urine of participants aged 25–47 with endometriosis who experience acyclic pelvic pain compared to samples from age-matched controls who did not report acyclic pelvic pain (P = 0.007) such that the presence of acyclic pelvic pain increased the NPV of EGF to 96.8% as determined by Bayes’ formula. These data demonstrate that urinary EGF has potential as a novel, non-invasive, accurate and objective biomarker for endometriosis diagnosis and prognosis of this disease.Supplementary InformationThe online version contains supplementary material available at 10.1186/s40364-026-00976-2.Keywords: iTRAQ, Biomarker, Non-invasive, Endometriosis, Epidermal Growth Factor, Urine To the Editor, Endometriosis is a chronic inflammatory estrogen-dependent condition marked by growth of endometrial-like tissue outside the uterus that affects approximately 10% of women [1]. Advances in imaging have improved diagnosis of certain subphenotypes of endometriosis, but superficial peritoneal lesions with a revised American Society for Reproductive Medicine (rASRM) stage of I or II and patients with negative imaging findings or failure with empirical treatment can only be definitively diagnosed via surgical visualization [2]. While imaging can play an important role in diagnosis, imaging alone cannot rule out disease [3, 4]. The lack of biomarkers of endometriosis leaves laparoscopy or laparotomy as the only option for objective diagnosis, contributing to the average 7-year delay from symptom onset to confirmed presence of endometriosis, leading to an average age of diagnosis in the fourth decade despite most patients experiencing symptoms before age 25 [5–7]. The identification of biomarkers to detect endometriosis could improve this timeline [8]. It would also provide insight into the pathophysiology of this disease, which could contribute to the development of personalized and more effective therapies. Urine is an ideal medium for non-invasive diagnostic and prognostic testing, rich with proteins and peptides that can serve as biomarkers. With a person’s average urine output of approximately 1.5 L per day, urine collection is simple, economical and non-invasive [9]. We set out to discover potential urinary biomarkers that could differentiate between those with and without endometriosis. We used an unbiased and quantitative proteomics discovery approach, using urine samples from an existing cohort sampled from participants with surgically-confirmed endometriosis (Supplemental Table 1). iTRAQ was used in the discovery phase to identify and quantitate the urinary proteome of women with endometriosis and controls. We used an 8-plex iTRAQ protocol as previously described [9]. Employing a conservative 2.0-fold cutoff for the differential expression of urinary proteins from participants with endometriosis compared to samples from sex- and age-matched controls, we identified 26 proteins that were downregulated and 30 proteins that were upregulated in the urine of participants with endometriosis (Fig. 1). Total protein quantitation was determined using the Bradford Assay as previously reported [10, 11]. Four proteins identified via iTRAQ were selected for validation based on physiological relevance and availability of reliable ELISA kits. In participants 25–47 years old, normalized EGF levels (pg/ug) were found to be significantly lower in the urine of endometriosis participants compared to controls (control n = 24; endometriosis n = 22; P = 0.02, Supplemental Table 2). Predictive accuracy of urinary EGF for diagnosis of endometriosis in this group was 0.701 (Fig. 2A). The other three analyzed proteins (RRBP1, MAFA and HPX) showed no ability to discriminate between participants with endometriosis and controls (data not shown) EGF ELISA results showed no significant difference in the concentration of EGF in urine between women with and without endometriosis when considering the whole cohort (Supplemental Table 2). A commonly reported symptom of endometriosis is acyclic pelvic pain [12]. Of those participants with endometriosis who reported the presence of acyclic pelvic pain, normalized urinary EGF levels were significantly lower than in the urine of controls who did not report the presence of acyclic pelvic pain (Supplemental Table 3). The accuracy of EGF in predicting the presence of endometriosis in women with acyclic pelvic pain is 0.767 (Fig. 2B). (A) ROC curve for urinary EGF (pg/ug) in participants aged 25-47 with endometriosis compared to age-matched controls (p-value=0.01). (B) ROC curve for urinary EGF (pg/ug) in participants ages 25-47 with endometriosis and acyclic pelvic pain compared to age-matched controls without acyclic pelvic pain (P<0.001). (C) Quartiles of urinary EGF levels (pg/ug) with associated predicted probability of endometriosis in women 25-47 years of age and (D) in those same women who also have acyclic pelvic pain. Predicted probability was calculated for each quartile using Bayes’ formula assuming a 10% prevalence of endometriosis Bayes’ formula was used with a population estimate of endometriosis prevalence of 10%, the prevalence of endometriosis in the general population, to determine PPV and NPV (Supplemental Table 4) [1]. A cutoff of < 800 pg/ug resulted in a negative predictive value of 94.3%, sensitivity of 64% and specificity of 67% in the urine of endometriosis participants vs. controls aged 25–47 years. In those same samples from endometriosis participants who also experienced acyclic pelvic pain compared to control samples, the same cutoff of < 800 pg/ug gave an even higher NPV of 96.8%, sensitivity of 80% and specificity of 68%. These NPVs suggest urinary EGF has promise as a clinical tool to rule out endometriosis in those with high levels, especially in the presence of acyclic pelvic pain. In addition to a binary cutoff value, we determined the probability of endometriosis based on quartile ranges of urinary EGF levels (Fig. 2C and D). We found that women 25–47 years of age with low urinary EGF levels had approximately double the odds of endometriosis compared to those with higher urinary EGF levels (Fig. 2C). Furthermore, when the presence of acyclic pelvic pain is considered, the risk increases approximately three-fold for those women with low urinary EGF levels (Fig. 2D). Urine from participants with much higher urinary EGF levels (> 1200 pg/ug) had a 0–1% predicted probability of disease, further supporting the clinical potential of urinary EGF testing as a tool to rule out endometriosis. EGF shows promise as a potential novel, non-invasive biomarker for endometriosis that could enable early and frequent testing for this disease. Supplementary Information Below is the link to the electronic supplementary material. Supplementary Material 1 Abbreviations Isobaric tags for relative and absolute quantification Epidermal Growth Factor Enzyme-linked immunosorbent assays Negative predictive value Revised American Society for Reproductive Medicine Nonsteroidal anti-inflammatory drugs Boston Children’s Hospital Brigham and Women’s Hospital World Endometriosis Research Foundation Endometriosis Phenome and Biobanking Harmonization Project Ribosome-binding protein 1 Transcription factor MafA Hemopexin Interquartile range Receiver operating characteristic Area under the curve Positive predictive value Author contributions ERRG and CCD contributed to the design of the study, data curation, analysis and were major contributors in writing and revising the manuscript. KK contributed to the design of the study, data curation and analysis. SJS contributed to analysis and writing and revising the manuscript. DZ contributed to analysis, supervision and writing and revising the manuscript. AFV and ALS contributed resources and participated in manuscript review and editing. KLT contributed to manuscript review and editing. STD contributed to analysis. SAM contributed resources and participated in writing and revising the manuscript. MSR contributed to funding acquisition and manuscript review and editing. TAL contributed to analysis and manuscript review and editing. MAM contributed to the design of the study, data curation, analysis, funding acquisition, supervision and was a major contributor in writing and revising the manuscript. Funding The authors gratefully acknowledge the support of the Nile Albright Research Foundation (MAM), the Thea and James M. Stoneman Charitable Foundation (MAM) and the Marriott Daughters Foundation (MAM, MR, SAM, KLT, ALS, AFV). Financial support for establishment of and data and urine collection within the A2A cohort was provided by the J. Willard and Alice S. Marriott Foundation (SAM, KLT, ALS, AFV). None of these funders had a role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Data availability The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request. Declarations Ethics and consent to participate The study was approved by the Boston Children’s Hospital (BCH) Institutional Review Board (Assurance Identification #: FWA00002071) on behalf of both BCH and Brigham and Women’s Hospital (BWH). Samples of urine were collected and processed as part of the Women's Health Study: From Adolescence to Adulthood (A2A) cohort. This includes patients with a surgical diagnosis from BCH or BWH or patients receiving follow-up treatments at BCH or BWH who received a prior documented surgical diagnosis elsewhere. Controls were females without a known diagnosis of endometriosis. They were recruited through local clinics, local advertisements, online postings and/or word of mouth and were identified to ensure sampling from the communities served by these two hospitals and thus the underlying population. All participants could have been receiving standard of care for any medical conditions including pelvic pain and are thus representative of the underlying general population. Samples were collected following the World Endometriosis Research Foundation (WERF) Endometriosis Phenome and Biobanking Harmonization Project (EPHect) protocol. Written informed consent was obtained, with parental consent and participant assent for girls <18 years of age at enrollment. All procedures involving human participants were performed in accordance with the ethical standards of the U.S. Common Rule, the principles of the Belmont Report and the ethical tenets outlined in the Declaration of Helsinki. Consent for publication Not applicable. Competing interest The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References Associated Data Supplementary Materials Supplementary Material 1 Data Availability Statement The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.