{"paper_id":"1052bbc6-43ac-4ee5-a5b8-9b07501ab7b9","body_text":"Oncogenic Mutations in Eutopic Endometrium \n \n 1\nOncogenic Mutations in Histologically Normal Endometrium: The \nNew Normal? \n \nVivian Lac1, Tayyebeh M. Nazeran1,2, Basile Tessier-Cloutier3, Rosalia Aguirre-Hernandez4, \nArianne Albert5, Amy Lum1, Jaswinder Khattra4, Teresa Praetorius6, Madeline Mason1, Derek \nChiu1, Martin Köbel7, Paul J. Yong8,9, Blake C. Gilks2,3, Michael S. Anglesio3,8,*, David G. \nHuntsman1,3,4,8,* \n \nAffiliations: \n1. Department of Molecular Oncology, BC Cancer Research Centre, Room 3-218, 675 West \n10th Ave, Vancouver, BC V5Z 1L3, Canada \n2. Department of Anatomical Pathology, Vancouver General Hospital, 899 W 12th Ave, \nVancouver, BC V5Z 1M9, Canada \n3. Department of Pathology and Laboratory Medicine, Rm. G227 – 2211 Wesbrook Mall, \nUniversity of British Columbia, Vancouver, BC V6T 2B5, Canada \n4. Contextual Genomics, 2389 Health Sciences Mall #204, Vancouver, BC V6T 1Z3, Canada \n5. Women’s Health Research Institute, 4500 Oak St, Vancouver, BC V6H 2N9, Canada \n6. Department of Women’s Health, Tuebingen University Hospital, Calwerstrasse 7, 72076 \nTuebingen, Germany \n7. Department of Pathology and Laboratory Medicine, University of Calgary, 2500 University \nDr. NW, Calgary, AB T2N 1N4, Canada \n8. Department of Obstetrics and Gynaecology, University of British Columbia, Suite 930, 1125 \nHowe Street, Vancouver, BC V6Z 2K8, Canada \n9. BC Women’s Centre for Pelvic Pain & Endometriosis, BC Women’s Hospital and Health \nCentre, Women' Health Centre, F2 - 4500 Oak St., Vancouver, BC V6H 3N1, Canada \n \nCorresponding Authors: \nDavid G. Huntsman, Department of Molecular Oncology, BC Cancer Research Centre, 675 West \n10th Ave, Vancouver, BC V5Z 1L3, Canada. Phone: 604-675-8211, Email: \ndhuntsma@bccancer.bc.ca\n \n \nMichael S. Anglesio, Department of Molecular Oncology, BC Cancer Research Centre, 675 \nWest 10th Ave, Vancouver, BC V5Z 1L3, Canada. Phone: 604-675-8211, E-mail: \nm.anglesio@ubc.ca \n \nConflicts of Interest: \nDGH is a co-founder and shareholder of Contextual Genomics Inc., a for profit company that \nprovides genomic diagnostics and reporting to assist in cancer patient treatment. RA-H and JK \nhave a patent MOLECULAR QUALITY ASSURANCE METHODS FOR USE IN \nSEQUENCING pending and are currently employed by Contextual Genomics Inc. The \nremaining authors have no conflicts of interest to declare. \n \n  \n.CC-BY-NC-ND 4.0 International licenseunder a \nnot certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available \nThe copyright holder for this preprint (which wasthis version posted February 26, 2019. ; https://doi.org/10.1101/561050doi: bioRxiv preprint \n\nOncogenic Mutations in Eutopic Endometrium \n \n 2\nAbstract:  \nThe presence of somatic driver mutations in endometriosis has previously been believed to \nrepresent early events in transformation, however our group and others have described such \nmutations in roughly one-third of cases of deep infiltrating or iatrogenic endometriosis. These \nforms of endometriosis rarely progress to malignancy. Recent studies have also shown somatic \ndriver mutations in normal skin, blood, peritoneal washings, and esophageal epithelium. Such \nfindings prompt speculation on whether such mutations exist in the eutopic endometrium – the \nlikely tissue of origin of endometriosis. In the current study we investigated the presence of \nsomatic driver mutations in histologically normal endometrium from women lacking evidence of \ngynecologic malignancy or endometrial hyperplasia. Twenty-five women who underwent \nhysterectomies and 85 women who underwent endometrial biopsies were included in this study. \nFormalin-fixed, paraffin-embedded tissue specimens were analyzed by means of targeted \nsequencing followed by orthogonal validation with droplet digital PCR. PTEN and ARID1A \nimmunohistochemistry (IHC) was also performed as surrogates for inactivating mutations in the \nrespective genes. Overall, we observed somatic driver-like events in over 50% of histologically \nnormal endometrial samples analyzed, which included hotspot mutations in KRAS, PIK3CA, and \nFGFR2 as well as PTEN-loss by IHC. Analysis of anterior and posterior samplings collected \nfrom women who underwent hysterectomies is consistent with the presence of somatic driver \nmutations within clonal pockets spread throughout the uterus. The prevalence of such oncogenic \nmutations also increased with age (OR: 1.05 (95% CI: 1.00 – 1.10), p = 0.035). These findings \nhave implications on our understanding of aging and so-called “normal tissues”, thereby \nnecessitating caution in the utilization of mutation-based early detection tools for endometrial or \nother cancers. \n \nKeywords: somatic mutations; normal tissue; aging; endometrium; digital PCR\n \n.CC-BY-NC-ND 4.0 International licenseunder a \nnot certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available \nThe copyright holder for this preprint (which wasthis version posted February 26, 2019. ; https://doi.org/10.1101/561050doi: bioRxiv preprint \n\nOncogenic Mutations in Eutopic Endometrium \n \n 3\nIntroduction: \nEndometriosis is a chronic disease affecting approximately 10 percent of reproductive-aged \nwomen. It leads to a variety of clinical symptoms including chronic (often cyclical) pain, \ndyspareunia, dysmenorrhea, and infertility[1,2]. Endometriosis is the major risk factor for clear \ncell and endometrioid carcinomas of the ovary (endometriosis-associated ovarian cancers; \nEAOCs) [3-5]. Observations of endometriotic lesions (consisting of endometrial-like glands and \nstroma) contiguous with EAOCs combined with molecular findings from several sequencing \nstudies have further established endometriosis as the precursor of EAOCs[6-8]. In these studies, \noncogenic mutations in genes such as ARID1A and PIK3CA[7,8] have been traced back to \nendometriotic lesions within the same patients, thereby implicating such alterations as early \nevents towards malignant transformation. However, we have found somatic driver mutations in \napproximately one-third of deep infiltrating endometriosis and iatrogenic endometriosis cases \n(forms of endometriosis with little malignant potential)[9,10]. These findings suggest that such \nmutations may be an inherent feature of endometriosis itself, perhaps conferring survival \nadvantages that allow lesions to persist ectopically. However, in order to study this hypothesis, it \nis important to clarify whether uterine epithelial mutations can pre-exist endometriosis and be \npresent in the cells prior to implantation at an ectopic site. \n \nIn the current study, we sought to explore the mutational landscape of the eutopic (uterine) \nendometrium from women lacking evidence of gynecologic malignancy by means of targeted \nsequencing and immunohistochemistry (IHC) studies. Although the origins of endometriosis \nremain contentious, the concept of retrograde menstruation stands as a leading theory. Herein, \nendometrial fragments (originating from the uterus) are refluxed upwards through the fallopian \ntubes and are implanted within the pelvic cavity[11]. This theory is supported by several lines of \nevidence including the high prevalence of endometriosis in females with congenital outflow \nobstruction[12] and the successful generation of endometriotic lesions in primate models of \nendometriosis via intrapelvic injection of menstrual endometrium[13] or cervical occlusion[14]. \nConsidering the dynamic nature of the endometrium as it undergoes many cycles of regeneration \nand tissue breakdown throughout a woman’s reproductive years[15] as well as the recent \ndescription of somatic mutations in normal tissues including the skin[16], blood[17], peritoneal \nwashings[18], and the esophageal epithelium[19], it is plausible that somatic mutations may pre-\nexist in the eutopic endometrium. Moreover, a series of previous studies have described “latent \nprecancers” (alterations in otherwise normal tissue with unknown contribution to long-term \ncancer risk) in the form of PTEN-null and PAX2-null glands within histologically normal \nendometrium[20-22]. Application of next-generation sequencing techniques will broaden our \nunderstanding of latent precancers, or pre-existing somatic alterations, in the eutopic \nendometrium. This is particularly important since these tissues often serves as a reference for \nmolecular studies on endometriosis, endometrial cancer, and EAOCs.  \n.CC-BY-NC-ND 4.0 International licenseunder a \nnot certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available \nThe copyright holder for this preprint (which wasthis version posted February 26, 2019. ; https://doi.org/10.1101/561050doi: bioRxiv preprint \n\nOncogenic Mutations in Eutopic Endometrium \n \n 4\nMaterials and Methods: \nPatient Identification and Specimen Collection \nWe obtained formalin-fixed and paraffin-embedded (FFPE) endometrial tissue specimens from \n25 women who underwent hysterectomies (Supplementary Table S1) and 85 women who \nunderwent either dilation and curettage (D&C) or endometrial biopsies (Supplementary Table \nS2). As confirmed by pathologist review, all women lacked evidence of gynecologic malignancy \nor endometrial hyperplasia and all tissue specimens were obtained from the Department of \nAnatomical Pathology at the Vancouver General Hospital in Vancouver, Canada. For \nhysterectomy patients (Hx cohort), we collected and subsequently analyzed two blocks to \nrepresent endometrial tissue from the posterior uterus and anterior uterus from each patient. For \nD&C or endometrial biopsy patients (Bx cohort), we targeted collection of endometrial tissue \nspecimens to represent women of various ages (ranging from 20 – 61 years of age). Specimen, \ndata collection and experiments were approved by the UBC BC Cancer Agency Research Ethics \nBoard [H05-60119, H02-61375 and H08-01411]. \n \nSample Processing and DNA Extraction \nHysterectomy specimens were sectioned at 8µm onto glass slides, deparaffinized with xylene \nand stained with 10% diluted hematoxylin and eosin (H&E). Specimens were manually \nmacrodissected to enrich the epithelial layer using the tip of a 20-guage needle under a stereo \nmicroscope. DNA from hysterectomy specimens was extracted using the ARCTURUS® \nPicoPure® DNA Extraction Kit (ThermoFisher Scientific, USA). D&C and endometrial biopsy \nspecimens were cut into 3 – 6 10µm tissue scrolls. DNA from D&C and endometrial biopsy \nspecimens was extracted using the QIAamp® DNA FFPE Tissue Kit (QIAGEN, USA). All \nDNA was quantitated using the Qubit 2.0 Fluorometer (ThermoFisher Scientific, USA). \n \nTargeted Sequencing \nAll specimens were sequenced using FIND IT\nTM version 3.4 (Contextual Genomics, Canada), a \nproprietary hypersensitive cancer hotspot assay which covers over 120 hotspots and 17 exons \nspanning 33 genes (Supplementary Table S3)[23]. Libraries were constructed using 75ng of total \nDNA input. Candidate somatic variants for orthogonal validation were selected from defined \ncancer hotspots reported in the Catalogue of Somatic Mutations in Cancer (COSMIC)[24], with \nprobability scores \n≥  0.8 (indicating the likelihood that a variant belongs to the mutation class as \nopposed to the artifact class) and variant allele frequency (VAF) ≥  0.8%. \n \nOrthogonal Validation via ddPCR \nDroplet digital polymerase-chain-reaction (ddPCR) was used to orthogonally validate hotspot \nmutations. DNA from each specimen was pre-amplified for 10 cycles before subsequent droplet \ngeneration using the QX200 Droplet Generator (Bio-Rad Laboratories, USA)[25]. In most cases, \nthe same aliquot of extracted DNA used for targeted sequencing was also used for ddPCR \nvalidation (Supplementary Table S5 and S6). After thermal cycling, the QX200 Droplet Reader \n(Bio-Rad Laboratories, USA) was used to quantify droplets. See Supplementary Table S7 for a \nlist of primers used and Supplementary Table S8 for ddPCR assay conditions. \n \nPTEN and ARID1A Immunohistochemistry \nLoss of PTEN immunoreactivity was used as a surrogate for PTEN loss-of-function mutations. \nFollowing the protocol outlined in our previous study[10], specimens were stained on the \n.CC-BY-NC-ND 4.0 International licenseunder a \nnot certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available \nThe copyright holder for this preprint (which wasthis version posted February 26, 2019. ; https://doi.org/10.1101/561050doi: bioRxiv preprint \n\nOncogenic Mutations in Eutopic Endometrium \n \n 5\nVentana Discovery Ultra (Ventana Medical Systems, USA) immunostainer using a 1:25 dilution \nof rabbit monoclonal antibody, 138G6 (Cell Signaling, USA). Similarly, loss of nuclear ARID1A \nimmunoreactivity was used as a surrogate for ARID1A loss-of-function mutations[10,26,27]. \nSpecimens were stained on the Dako Omnis (Agilent Technologies, USA) automated \nimmunostainer using a 1:2000 dilution of ARID1A rabbit monoclonal antibody EPR13501 \n(ab182560, Abcam, USA). TMN and BT-C. scored all ARID1A and PTEN immunostained \nslides. \n \nResults: \nSomatic Driver Mutation Frequencies in Normal Endometrium \nAs shown in Figure 1, we found that 16 of 25 women (64%) who underwent hysterectomy and \n43 of 85 women (51%) who underwent D&C or endometrial biopsy harbored at least one \nsomatic alteration within their endometrial tissue samplings. Given our sample sizes, the \nprevalence of these alterations (whether specifically or overall) did not significantly differ \nbetween Hx and Bx cohorts (Supplementary Table S9). The most common somatic alterations \nwere hotspot KRAS and PIK3CA mutations (affecting 31 of 110 and 14 of 110 total patients \nrespectively), and focal loss of PTEN expression (defined by absence of staining in a small \ncluster of glands among normally expressing glands) (30 of 110 total patients) (Supplementary \nFigure S1). No patients exhibited loss of ARID1A either focally or globally. \n \nAnatomical Distribution of Oncogenic Changes \nFor each woman in the Hx cohort, we independently collected and sequenced DNA from two \ntissue blocks representing endometrial tissue from the anterior and posterior uterus within the \nsame patient. Somatic hotspot mutations were either completely absent (15 of 25 patients) or \ndissimilar (9 of 25 patients) between anterior and posterior samplings (Supplementary Figure \nS2). A single patient (Hx_25) harbored a KRAS G12A mutation in both samplings, however \nanatomical distortion of the uterus in this patient resulted in an inability to determine the spatial \ndistinction between the two blocks of endometrial tissue (Table 1). \n \nDriver Mutations Across Age \nAccumulation of somatic alteration with increasing age has previously been suggested even in \nthe context of normal tissues[16-19]. To assess the relationship between age and presence of \nsomatic cancer-driver alterations, we generated a logistic regression model based on the \nsequencing and IHC findings for the 25 women in the Hx cohort and 85 women in the Bx cohort \n(Figure 2). Based on our model, the likelihood of a woman harbouring a somatic alteration in her \nendometrial tissue increases by 5% per year (OR = 1.05, 95% CI = 1.00 – 1.10, p = 0.035, Wald \ntest). Moreover, a model incorporating both age and menstrual phase (proliferative versus \nsecretory phase) indicates that the prevalence of mutation is independent of menstrual phase (p = \n0.6309, Wald test) (Supplementary Figure S3). \n \nComparison of Normal Endometrium Findings with Endometrial Cancer \nWe compared our data with endometrial cancer data from The Cancer Genome Atlas \n(TCGA)[28] to better interpret mutational findings within the normal endometrium. Because our \nanalysis is restricted to the hotspot regions in the 33 genes covered by the FIND IT\nTM assay \n(Supplemental Table S3), we limited our comparison to oncogenic mutations within these \nregions for the TCGA dataset. For our normal endometrial tissue cohort, we pooled our data \n.CC-BY-NC-ND 4.0 International licenseunder a \nnot certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available \nThe copyright holder for this preprint (which wasthis version posted February 26, 2019. ; https://doi.org/10.1101/561050doi: bioRxiv preprint \n\nOncogenic Mutations in Eutopic Endometrium \n \n 6\nfrom both the Hx cohort (n=25) and Bx cohort (n=85). Overall, endometrial cancer samples \nmore frequently harbored oncogenic mutations compared to the patients examined in our study \n(96% vs 54%, p < 0.0001, Student’s t-test) (Figure 3A; Supplementary Table S10). Women with \nendometrial cancer were far more likely to harbor a somatic mutation at a given age (OR: 10.8, \n95% CI: 3.83 – 33.3, p = 1.5 e-05, Wald test) (Supplementary Figure S4). \n \nGene-specific analysis mutation frequencies were either indistinguishable or higher in cancer \nspecimens, except for KRAS driver mutations. We found a distinct trend (p = 0.0728) suggesting \na greater proportion of non-cancer patients harbored KRAS driver mutations in their endometrial \ntissue, 31 of 110 (28%), compared with endometrial cancer patients, 48 of 248 (19%) (Figure \n3A; Supplementary Table S10). Indeed, in a model that incorporates age and KRAS mutation, \nwhilst mutation rates in normal samples clearly increase with age (OR = 1.06, 95% CI = 1.01 – \n1.11, p = 0.01953, Wald test), in the cancer samples the opposite trend is seen (OR = 0.96, 95% \nCI = 0.94 – 0.99, p = 0.0164, Wald test) (Figure 3B). \n \nDiscussion: \nIn many cancers, tumor progression is largely dependent on the accumulation of so-called \n“driver mutations” that confer selective growth advantages on affected subclones compared to \nsurrounding cells[29-31]. EAOCs, in fact are characterized by mutations affecting PIK3CA, \nPPP2R1A, ARID1A, KRAS, CTNNB1, and PTEN[32]. Sporadic endometrial cancers also have a \nsimilar driver mutation landscape characterized by mutations affecting PIK3CA, PIK3R1, \nARID1A, KRAS, CTNNB1, PTEN, FGFR2, and alterations in mismatch repair proteins[33]. The \neutopic endometrium, an extremely dynamic tissue over the course of a woman’s reproductive \nyears, is both the tissue of origin for endometrial cancer and the presumed origin of EAOCs[6-8]. \nTherefore, it serves as an important reference for the studies of these cancers as well as \nendometriosis. In the current study, we have described (with orthogonal validation) the presence \nof somatic, driver mutations within histologically unremarkable, eutopic endometrium of over \nhalf of women lacking evidence of gynecologic malignancy. \n \nOur findings are consistent with oncogenic mutations existing in small clonal pockets throughout \nthe endometrium. Amongst informative patients, oncogenic mutations were dissimilar between \nthe anterior and posterior samplings (i.e. if a particular point mutation were found in one \nsampling, it was not detected in the other sampling) – these findings were confirmed by ddPCR \ntesting of both samplings for any mutation found in either block (Table 1; Supplementary Table \nS5). The VAFs of reported somatic mutations were generally low (between 0.8 – 11.9% in \nhysterectomy or D&C/biopsy cases – see Table 1; Supplementary Table S5, S6), despite some \nenrichment of the glandular epithelial layer. This suggests that oncogenic mutations may be \ngland-level restricted, affecting only single (or a small number of) endometrial glands across a \nbreadth of many glands from a given tissue specimen. This hypothesis seems well supported by \ntwo recent studies wherein individual endometrial glands were laser-captured from different \nareas within the endometrium of the same patient. Both studies suggest that different glands \nharbored different mutations[34,35]. These studies and our own results are also consistent with \nthe findings by Mutter et al. (2014)[20],where PTEN-loss appears zonally in single, or small \nclusters of, glands (Supplementary Figure S1). Therefore, although our study represents broader \nsampling of the endometrium and mutations, our observations and those of others[34,35] are \nconsistent with driver mutations, such as KRAS G12 mutations, occurring in small, clonal \n.CC-BY-NC-ND 4.0 International licenseunder a \nnot certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available \nThe copyright holder for this preprint (which wasthis version posted February 26, 2019. ; https://doi.org/10.1101/561050doi: bioRxiv preprint \n\nOncogenic Mutations in Eutopic Endometrium \n \n 7\npockets throughout the uterus. Furthermore, it seems likely that acquisition of these mutations \nprovides sufficient selective advantages to expand within an entire gland of affected endometrial \ncells, allowing this population to be detectable by our methods. \n \nSomatic mutations have been previously described to accumulate with age in a variety of non-\ncancerous tissues[18,19,36] and thus we sought to determine the relationship between age and \nprevalence of somatic mutation in the normal endometrium. Overall, we found that the \nprevalence of oncogenic mutations in normal endometrium indeed increases by age; the risk of \nharboring such a mutation increases by approximately 5% per year. Moreover, the influence of \nage on mutation prevalence is independent of the phase of the menstrual cycle (Supplementary \nFigure S3). Comparing our data to the TCGA endometrial cancer dataset reveals that the \nprevalence of driver mutations is higher across age in endometrial cancer (or comparably as \nuncommon) in all genes except for KRAS. In our study, KRAS driver mutations occurred in over \none-quarter of eutopic endometrium cases, illustrating a trend to greater prevalence in non-\ncancerous vs endometrial cancer samples. Although the KRAS mutation rates in normal \nendometrial samplings increased with age the opposite was seen in the TCGA cancer samples. \nThe latter observation is likely due to distinct KRAS mutation rates in endometrial cancer \nsubtypes that, themselves, have distinct age distributions[28].  Since patients in our study \nunderwent hysterectomies, D&Cs, or endometrial biopsies for benign uterine-related pathologies \nor related concerns, it is plausible that KRAS mutations may play a role in such benign/non-\ncancer pathologies such as those described in arteriovenous malformations of the brain[37] or as \nwe have previously observed in endometriosis[9,10]. It is also possible that early activation of \nKRAS induces senescence[38] in endometrial cells, thereby inhibiting progression towards \nmalignant transformation. In contrast to our high frequency KRAS mutations, CTNNB1 and \nARID1A are commonly mutated in endometrial cancer and EAOCs[32,33] but were not observed \nin any of our normal endometrium specimens. This may suggest that, within the context of \nendometrial-derived carcinomas, such mutations are later events in tumorigenesis or that \naccumulation of mutations in cancer-destined precursors is etiologically distinct from age-related \naccumulations. \n \nLastly, it is important to note that many of the cancer specimens in the TCGA dataset harbor \nmultiple driver mutations within the same tumor. Although some of our normal endometrium \nspecimens harbored multiple driver alterations within the same block, this was relatively \nuncommon and it remains unclear if mutations affected overlapping populations or independent \nsubclones. The recent study by Moore et al. found two or more driver mutations in single \nendometrial glands in 22% of specimens[35], suggesting an elevated probability, albeit still non-\nconclusive, that driver mutations may co-exist in the same non-cancerous cells. \n \nWithin the eutopic endometrium, detection of driver mutations increases with age despite the \nlack of remarkable histological changes. Other factors, particularly oral contraceptive use or \nhormonal intrauterine device use may further modulate (reduce) risk of accumulating driver \nmutations within endometrial tissues, such as has been observed as a decline in the frequency of \nPTEN-null glands in women using these contraceptive methods[21] . Because of the lack of \nfollow-up with patients in our study, it remains unclear what role, if any, these mutations play in \nthe eutopic endometrium. Other obvious questions include whether mutations persist, harbored \nin early progenitor populations, or do they disappear at each menstrual cycle re-emerging de-\n.CC-BY-NC-ND 4.0 International licenseunder a \nnot certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available \nThe copyright holder for this preprint (which wasthis version posted February 26, 2019. ; https://doi.org/10.1101/561050doi: bioRxiv preprint \n\nOncogenic Mutations in Eutopic Endometrium \n \n 8\nnovo with the rapid proliferation of the endometrial lining each month[20]?  Our observation of \nsurprisingly high frequency of driver mutations in the eutopic endometrium of women without \nevidence of malignancy or even subtle pathology suggests that care must be taken when \ninterpreting detected mutations in non-cancer affected individuals. Ongoing preclinical \ndevelopment of mutation detection assays have suggested non-cancer affected populations are \nunaffected by somatic driver alterations[39],  many of which overlapped with our assays. \nHowever, closer inspection suggests at least some studies may not show a full picture with both \nsampling methodology and control population age-bias potentially contributing to the failure to \ndetect driver mutations in non-cancer populations. The development of screening methods for \noccult or pre-cancerous disease will need to incorporate appropriate specimen sampling, age-\nappropriate control populations, and a solid understanding of normal age-related somatic \nmutation accumulation. \n \nIn summary, our findings highlight the complexity of the eutopic endometrium and challenge our \ncurrent understanding of so-called “normal tissues”. As the mutations identified are obviously \noncogenic in other tissues, these findings highlight the need to consider cell context and \nmicroenvironment along with mutation in the development of cancer.  \n  \n.CC-BY-NC-ND 4.0 International licenseunder a \nnot certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available \nThe copyright holder for this preprint (which wasthis version posted February 26, 2019. ; https://doi.org/10.1101/561050doi: bioRxiv preprint \n\nOncogenic Mutations in Eutopic Endometrium \n \n 9\nAcknowledgements: \nThis research is funded by the Canadian Cancer Society (Impact grant #701603), Canadian \nInstitutes of Health Research (Foundation grant #154290), and the Janet D. Cottrelle Foundation \nthrough the BC Cancer Foundation. The VGH & UBC Hospital Foundation and the BC Cancer \nFoundation provided funding to OVCARE: BC’s Ovarian Cancer Research Team. A Frederick \nBanting and Charles Best Canada Graduate Scholarship from the Canadian Institutes of Health \nResearch provided support to V. Lac. The Janet D. Cottrelle Foundation Scholars fund provided \nsupport to M.S. Anglesio. The Dr. Chew Wei Memorial Professorship in Gynecologic Oncology \nand the Canada Research Chairs program (Research Chair in Molecular and Genomic Pathology) \nprovided support to D.G. Huntsman. The Michael Smith Foundation for Health Research Health \nProfessional-Investigator Award provided support to P.J. Yong. \n  \nWe would also like to acknowledge the staff at the Genetic Pathology Evaluation Centre \n(GPEC), the Department of Anatomical Pathology at the Vancouver General Hospital in \nVancouver, OVCARE’s gynaecological tissue bank, and the Calgary Laboratory Services for \ntheir assistance with specimen collection, immunohistochemical staining and optimization, and \ntissue sectioning. \n \n \nStatement of Author Contributions: \nVL, MSA, and DGH designed the study. TMN and BT-C identified specimens for the study, VL, \nAL, JK, TP, and MM processed samples, and performed experiments. TMN, BT-C, and MK \nreviewed pathology and generated/reviewed IHC data. VL and RA-H performed data analysis. \nAA and DC provided statistical design input and oversight. PJY and BCG provided clinical input \nto the study design. VL drafted the manuscript. All authors revised the manuscript and approved \nsubmission of the final version. \n  \n.CC-BY-NC-ND 4.0 International licenseunder a \nnot certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available \nThe copyright holder for this preprint (which wasthis version posted February 26, 2019. ; https://doi.org/10.1101/561050doi: bioRxiv preprint \n\nOncogenic Mutations in Eutopic Endometrium \n \n 10\nReferences: \n1. Vercellini P, Vigano P, Somigliana E, et al. Endometriosis: pathogenesis and treatment. Nat \nRev Endocrinol 2014;10:261-275. \n2. Giudice LC. Clinical practice. Endometriosis. N Engl J Med 2010;362:2389-2398. \n3. Sampson JA. Endometrial carcinoma of the ovary arising in endometrial tissue in that organ. \nAmerican Journal of Obstetrics & Gynecology 1925;9:111-114. \n4. Pearce CL, Templeman C, Rossing MA, et al. Association between endometriosis and risk of \nhistological subtypes of ovarian cancer: a pooled analysis of case-control studies. Lancet Oncol \n2012;13:385-394. \n5. Saavalainen L, Lassus H, But A, et al. Risk of Gynecologic Cancer According to the Type of \nEndometriosis. Obstet Gynecol 2018;131:1095-1102. \n6. Sato N, Tsunoda H, Nishida M, et al. Loss of heterozygosity on 10q23.3 and mutation of the \ntumor suppressor gene PTEN in benign endometrial cyst of the ovary: possible sequence \nprogression from benign endometrial cyst to endometrioid carcinoma and clear cell carcinoma of \nthe ovary. Cancer Res 2000;60:7052-7056. \n7. Wiegand KC, Shah SP, Al-Agha OM, et al. ARID1A mutations in endometriosis-associated \novarian carcinomas. N Engl J Med 2010;363:1532-1543. \n8. Anglesio MS, Bashashati A, Wang YK, et al. Multifocal endometriotic lesions associated with \ncancer are clonal and carry a high mutation burden. J Pathol 2015;236:201-209. \n9. Anglesio MS, Papadopoulos N, Ayhan A, et al. Cancer-Associated Mutations in \nEndometriosis without Cancer. N Engl J Med 2017;376:1835-1848. \n10. Lac V, Verhoef L, Aguirre-Hernandez R, et al. Iatrogenic endometriosis harbors somatic \ncancer-driver mutations. Hum Reprod 2019;34:69-78. \n11. Burney RO, Giudice LC. Pathogenesis and pathophysiology of endometriosis. Fertil Steril \n2012;98:511-519. \n12. Sanfilippo JS, Wakim NG, Schikler KN, et al. Endometriosis in association with uterine \nanomaly. Am J Obstet Gynecol 1986;154:39-43. \n13. D'Hooghe TM, Bambra CS, Raeymaekers BM, et al. Intrapelvic injection of menstrual \nendometrium causes endometriosis in baboons (Papio cynocephalus and Papio anubis). Am J \nObstet Gynecol 1995;173\n: 125-134. \n14. D'Hooghe TM, Bambra CS, Suleman MA, et al. Development of a model of retrograde \nmenstruation in baboons (Papio anubis). Fertil Steril 1994;62:635-638. \n15. Masuda H, Matsuzaki Y, Hiratsu E, et al. Stem cell-like properties of the endometrial side \npopulation: implication in endometrial regeneration. PLoS One 2010;5:e10387. \n16. Martincorena I, Roshan A, Gerstung M, et al. Tumor evolution. High burden and pervasive \npositive selection of somatic mutations in normal human skin. Science 2015;348:880-886. \n17. Jacobs KB, Yeager M, Zhou W, et al. Detectable clonal mosaicism and its relationship to \naging and cancer. Nat Genet 2012;44:651-658. \n18. Krimmel JD, Schmitt MW, Harrell MI, et al. Ultra-deep sequencing detects ovarian cancer \ncells in peritoneal fluid and reveals somatic TP53 mutations in noncancerous tissues. Proc Natl \nAcad Sci U S A 2016;113:6005-6010. \n19. Martincorena I, Fowler JC, Wabik A, et al. Somatic mutant clones colonize the human \nesophagus with age. Science 2018;362:911-917. \n20. Mutter GL, Monte NM, Neuberg D, et al. Emergence, involution, and progression to \ncarcinoma of mutant clones in normal endometrial tissues. Cancer Res 2014;74:2796-2802. \n.CC-BY-NC-ND 4.0 International licenseunder a \nnot certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available \nThe copyright holder for this preprint (which wasthis version posted February 26, 2019. ; https://doi.org/10.1101/561050doi: bioRxiv preprint \n\nOncogenic Mutations in Eutopic Endometrium \n \n 11\n21. Lin MC, Burkholder KA, Viswanathan AN, et al. Involution of latent endometrial precancers \nby hormonal and nonhormonal mechanisms. Cancer 2009;115:2111-2118. \n22. Monte NM, Webster KA, Neuberg D, et al. Joint loss of PAX2 and PTEN expression in \nendometrial precancers and cancer. Cancer Res 2010;70:6225-6232. \n23. Aparicio S, Shah S, Aguirre-Hernandez R, et al. MOLECULAR QUALITY ASSURANCE \nMETHODS FOR USE IN SEQUENCING. 2018;. \n24. Forbes SA, Beare D, Boutselakis H, et al. COSMIC: somatic cancer genetics at high-\nresolution. Nucleic Acids Res 2017;45:D783. \n25. Lac V, Huntsman DG. Distinct developmental trajectories of endometriotic epithelium and \nstroma: implications for the origins of endometriosis. J Pathol 2018;. \n26. Khalique S, Naidoo K, Attygalle AD, et al. Optimised ARID1A immunohistochemistry is an \naccurate predictor of ARID1A mutational status in gynaecological cancers. J Pathol Clin Res \n2018;. \n27. Kobel M, Anglesio MS, Brenton JD. You won't believe this old test ... that does cheap \nsingle-cell mutation detection. J Pathol Clin Res 2018;. \n28. Cancer Genome Atlas Research Network, Kandoth C, Schultz N, et al. Integrated genomic \ncharacterization of endometrial carcinoma. Nature 2013;497:67-73. \n29. Hanahan D, Weinberg RA. The hallmarks of cancer. Cell 2000;100:57-70. \n30. Hanahan D, Weinberg RA. Hallmarks of cancer: the next generation. Cell 2011;144:646-\n674. \n31. Stratton MR, Campbell PJ, Futreal PA. The cancer genome. Nature 2009;458:719-724. \n32. Worley MJ, Welch WR, Berkowitz RS, et al. Endometriosis-associated ovarian cancer: a \nreview of pathogenesis. Int J Mol Sci 2013;14:5367-5379. \n33. O'Hara AJ, Bell DW. The genomics and genetics of endometrial cancer. Adv Genomics \nGenet 2012;2012:33-47. \n34. Suda K, Nakaoka H, Yoshihara K, et al. Clonal Expansion and Diversification of Cancer-\nAssociated Mutations in Endometriosis and Normal Endometrium. Cell Rep 2018;24:1777-1789. \n35. Moore L, Leongamornlert D, Coorens TH, et al. The mutational landscape of normal human \nendometrial epithelium. bioRxiv 2018;. \n36. Risques RA, Kennedy SR. Aging and the rise of somatic cancer-associated mutations in \nnormal tissues. PLoS Genet 2018;14:e1\n007108. \n37. Nikolaev SI, Vetiska S, Bonilla X, et al. Somatic Activating KRAS Mutations in \nArteriovenous Malformations of the Brain. N Engl J Med 2018;378:250-261. \n38. Robinson G. The interrelationship of anxiety, hostility and guilt; some physiological \ncorrelations. J Asthma Res 1967;5:115-121. \n39. Wang Y, Li L, Douville C, et al. Evaluation of liquid from the Papanicolaou test and other \nliquid biopsies for the detection of endometrial and ovarian cancers. Sci Transl Med \n2018;10:10.1126/scitranslmed.aap8793. \n  \n.CC-BY-NC-ND 4.0 International licenseunder a \nnot certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available \nThe copyright holder for this preprint (which wasthis version posted February 26, 2019. ; https://doi.org/10.1101/561050doi: bioRxiv preprint \n\nOncogenic Mutations in Eutopic Endometrium \n \n 12\nTable 1. Mutational findings within normal endometrium samples from hysterectomy patients \nand corresponding variant allele frequencies (VAF) determine through targeted sequencing and \ndroplet digital PCR (ddPCR). \nPatient and \nBlock ID \nRegion of Uterus \nSampled \nPTEN IHC Driver Mutation \nIdentified \nVAF (%) - \nSequencing \nVAF (%) - \nddPCR \nHx_2A  ante rior  normal      \nHx_2B  posteri or  normal  ERBB2 S310F 0.974 0.792 \nHx_4A  ante rior  loss - zonal     \nHx_4B  posteri or  normal      \nHx_5A  ante rior  loss - zonal     \nHx_5B  posteri or  normal      \nHx_6A  ante rior  normal  KRAS G12V 2.654 1.63 \nHx_6B  posteri or  normal  KRAS G12A  1.408 1.08 \nHx_7A  posteri or  loss - zonal     \nHx_7B  ante rior  normal      \nHx_8A  posteri or  loss - zonal     \nHx_8B  ante rior  loss - zonal     \nHx_9A  ante rior  normal      \nHx_9B  posteri or  loss - zonal KRAS G13D 2.16 1.95 \nHx_10A  n/a normal      \nHx_10B  n/a loss - zonal     \nHx_12A  ante rior  normal  PIK3CA R88Q 1.337 1.23 \nHx_12B  posteri or  normal      \nHx_13A  posteri or  normal  PIK3CA H1047R 2.497 3.36 \nKRAS G12C 1.036 0.538 \nHx_13B  ante rior  normal      \nHx_16A  ante rior  normal  KRAS G12V 1.079 1.06 \nPIK3CA E542K  1.015 0.897 \nHx_16B  posteri or  loss - zonal PIK3CA H1047R 2.402 1.92 \nHx_18A  ante rior  normal  FGFR2 K659E 5.505 7.77 \nKRAS G12V 1.551 1.24 \nHx_18B  posteri or  loss - zonal     \nHx_21A  ante rior  normal  KRAS G12D 5.587 4.71 \nPIK3CA M1043I 2.39 2.27 \nHx_21B  posteri or  normal  KRAS G12V 6.112 6.32 \nPIK3CA H1047R 2.373 2.76 \nFGFR2 S252W  1.889 1.17 \nHx_22A  ante rior  loss - zonal     \nHx_22B  posteri or  normal      \nHx_23A  ante rior  n/a     \nHx_23B  posteri or  normal  KRAS G12D 11.866 10.2 \nHx_25A  n/a normal  KRAS G12A  1.584 1.60 \nHx_25B  n/a loss - zonal KRAS G12A  1.670 1.00 \n \n*Data for nine patients without detectable somatic mutations or abnormal immunohistochemistry \nfindings are omitted from this table. \n.CC-BY-NC-ND 4.0 International licenseunder a \nnot certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available \nThe copyright holder for this preprint (which wasthis version posted February 26, 2019. ; https://doi.org/10.1101/561050doi: bioRxiv preprint \n\nOncogenic Mutations in Eutopic Endometrium \n \n 13\nFigure Legends: \nFigure 1. Overview of somatic, oncogenic alterations detected in the eutopic endometrium from \n(A) hysterectomy patients or (B) dilation and curettage or endometrial biopsy patients. Red \nrectangles denote driver mutations detected by targeted sequencing, whereas blue rectangles \ndenote (zonal) loss detected through immunohistochemistry studies. The asterisk (*) denotes a \npatient wherein zonal loss of PTEN as well as a PTEN driver mutation was detected. \n \nFigure 2. The prevalence of somatic driver mutations in the eutopic endometrium across age. \nAccording to this logistic regression model, the prevalence of mutation increases with age (OR: \n1.05 (95% CI: 1.00 – 1.10), p = 0.035, Wald test). \n \nFigure 3. (A) Comparison of the prevalence of somatic mutations in endometrial cancer versus \nhistologically normal eutopic endometrium (mutations found in endometrial cancer but were not \ndetected in any eutopic endometrium sample are not displayed). (B) KRAS mutation prevalence \nacross age in endometrial cancer samples versus normal endometrium samples. In normal \nendometrium samples, the prevalence of KRAS mutation increases with age (OR = 1.06, 95% CI \n= 1.01 – 1.11, p = 0.01953, Wald test), whereas it decreases with age in endometrial cancer \nsamples (OR = 0.96, 95% CI = 0.94 – 0.99, p = 0.0164, Wald test). \n \nSupplementary Figure Legends: \nSupplementary Figure S1. PTEN IHC studies of eutopic endometrium specimens showing \nzonal loss of PTEN in epithelial cells and matching H&E staining. \n \nSupplementary Figure S2. Comparison of point mutations detected in anterior and posterior \nsamplings from patients in the Hx cohort. \n \nSupplementary Figure S3. Logistic regression model of mutation prevalence across age in \nsecretory phase versus proliferative phase endometrium \n \nSupplementary Figure S4. Logistic regression model of mutation prevalence across age in \nendometrial cancer versus normal endometrium. \n.CC-BY-NC-ND 4.0 International licenseunder a \nnot certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available \nThe copyright holder for this preprint (which wasthis version posted February 26, 2019. ; https://doi.org/10.1101/561050doi: bioRxiv preprint \n\n.CC-BY-NC-ND 4.0 International licenseunder a \nnot certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available \nThe copyright holder for this preprint (which wasthis version posted February 26, 2019. ; https://doi.org/10.1101/561050doi: bioRxiv preprint \n\n.CC-BY-NC-ND 4.0 International licenseunder a \nnot certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available \nThe copyright holder for this preprint (which wasthis version posted February 26, 2019. ; https://doi.org/10.1101/561050doi: bioRxiv preprint \n\n.CC-BY-NC-ND 4.0 International licenseunder a \nnot certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available \nThe copyright holder for this preprint (which wasthis version posted February 26, 2019. ; https://doi.org/10.1101/561050doi: bioRxiv preprint \n\n.CC-BY-NC-ND 4.0 International licenseunder a \nnot certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available \nThe copyright holder for this preprint (which wasthis version posted February 26, 2019. ; https://doi.org/10.1101/561050doi: bioRxiv preprint \n\n.CC-BY-NC-ND 4.0 International licenseunder a \nnot certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available \nThe copyright holder for this preprint (which wasthis version posted February 26, 2019. ; https://doi.org/10.1101/561050doi: bioRxiv preprint \n\n.CC-BY-NC-ND 4.0 International licenseunder a \nnot certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available \nThe copyright holder for this preprint (which wasthis version posted February 26, 2019. ; https://doi.org/10.1101/561050doi: bioRxiv preprint \n\n.CC-BY-NC-ND 4.0 International licenseunder a \nnot certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available \nThe copyright holder for this preprint (which wasthis version posted February 26, 2019. ; https://doi.org/10.1101/561050doi: bioRxiv preprint","source_license":"CC0","license_restricted":false}