hum. reprod.-2006-tokushige-782-7

Human Reproduction Vol.21, No.3 pp. 782–787, 2006 doi:10.1093/humrep/dei368

Advance Access publication October 27, 2005.

782 © The Author 2005. Published by Oxford University Press on behalf of the European Society of Human Reproduction and Embryology. All rights reserved.For Permissions, please email: [email protected]

High density of small nerve fibres in the functional layer of the endometrium in women with endometriosis

N.Tokushige1,3, R.Markham1, P.Russell2 and I.S.Fraser1

1Department of Obstetrics and Gynaecology, Queen Elizabeth II Research Institute for Mothers and Infants and 2Department of Pathology, University of Sydney, Sydney, 2006, Australia3To whom correspondence should be addressed. E-mail: [email protected]

BACKGROUND: Endometriosis is a common gynaecological disease and is frequently associated with recurrentand serious pelvic pain such as dysmenorrhoea and dyspareunia, but the mechanisms by which these symptoms aregenerated are not well understood. METHODS: Histological sections of endometrial tissue were prepared fromendometrial curettings and hysterectomies performed on women with endometriosis (n=25 and n=10, respectively) andwithout endometriosis (n=47 and n=35, respectively). These were stained immunohistochemically for the highly spe-cific polyclonal rabbit anti-protein gene product 9.5 (PGP9.5) and monoclonal mouse anti-neurofilament protein (NF)to demonstrate both myelinated and unmyelinated nerve fibres. RESULTS: Small nerve fibres were identifiedthroughout the basal and functional layers of the endometrium in all endometriosis patients, but were not seen in thefunctional layer of the endometrium in any of the women without endometriosis (P< 0.001). NF-immunoreactive nervefibres were present in the basal layer in all endometriosis patients but not in non-endometriosis patients, with oneexception (P< 0.001). CONCLUSIONS: Small nerve fibres detected in the functional layer in all women with endome-triosis may have important implications for understanding the generation of pain in these patients. The presence ofnerve fibres in an endometrial biopsy may be a novel surrogate marker of clinical endometriosis.

Key words: endometriosis/endometrium/immunohistochemistry/nerve fibres/pain

Introduction

Endometriosis, defined by the presence of focal endometrium-like tissue in sites outside the uterus (Binkovitz et al., 1991),is a gynaecological condition which is being increasingly rec-ognized in most societies, especially in developed countries.Frequently it is associated with recurrent and disabling symp-toms, primarily pelvic pain, yet little is understood about themechanisms by which symptoms are generated (Anaf et al.,2000). Investigators have studied the presence of nerve fibresin endometriotic plaques and, although there is some evidencefor the ingrowth of small nerve fibres into endometrioticplaques (Tamburro et al., 2003; Berkley et al., 2004), no neu-rofilament protein (NF)-immunoreactive nerve fibres could beidentified in the capsule of ovarian endometriomas (Al-Frozanet al., 2004).

The multiplicity of molecular and cellular abnormalitiesdemonstrated in the eutopic endometrium of women withendometriosis (Lessey et al., 1994; Ota and Tanaka, 1997;Healy et al., 1998) suggests that primary abnormalities in thistissue may have a role in the genesis of endometriosis and per-haps even in the genesis of symptoms. We speculated that thismight include local growth factors or cytokines for nerve fibreswhich, if also present in the ectopic endometrial-type tissue ofpatients with endometriosis, could induce the growth of nerve

fibres into the endometriotic foci and thus provide a mecha-nism for lesion-specific pain. We therefore studied the pres-ence of nerve fibres by immunohistochemistry in the basal andfunctional layers of the endometrium from women with andwithout visually and biopsy-proven endometriosis using highlyspecific markers polyclonal rabbit anti-protein gene product9.5 (PGP9.5) (Lundberg et al., 1988) and NF (Schlaepfer,1987) for both myelinated and unmyelinated nerve fibres.PGP9.5 is a highly specific pan-neuronal marker and it stainsboth myelinated and unmyelinated nerve fibres (Aα, Aβ, Aγ,Aδ, B and C fibres). NF is a highly specific marker for myeli-nated nerve fibres and it stains Aα, Aβ, Aγ, Aδ and B fibres.Aδ fibres are small myelinated fibres and transmit sharp, prick-ing localized pain to the central nervous system. C fibres aresmall unmyelinated fibres and transmit dull, aching, burningpoorly localized pain.

Materials and methods

Collection of tissue

This study was approved by the Human Ethics Committees of theCentral Sydney Area Health Service and the University of Sydney,and all women gave their informed consent for participation.

Endometrial tissue (uterine curettage) was collected from 25women with endometriosis (mean age, 27.5 years; range, 20–35) and

by guest on March 30, 2015

http://humrep.oxfordjournals.org/

Dow

nloaded from


Endometrial nerve fibres in women with endometriosis

783

47 women without endometriosis (mean age 32.7 years; range, 28–38)who underwent laparoscopy combined with hysteroscopy anddiagnostic curettage. The endometriosis patients all complained ofdysmenorrhoea and a range of related pain symptoms. The 47endometrial curettage tissue samples collected from women withoutlaparoscopic evidence of endometriosis were from patients undergo-ing tubal sterilization, assessment prior to tubal reanastomosis orinvestigation of infertility. Full-thickness uterine blocks, whichincluded the endometrium and myometrium, were selected from 10women with endometriosis (mean age 44.0 years; range, 42–46) and35 women without endometriosis (mean age 46.0 years; range, 38–54)who had undergone hysterectomy. The indications for hysterectomyin the 35 women without endometriosis included multiple small uter-ine fibroids, uterine prolapse, abdominal adhesions or abnormallyheavy menstrual bleeding.

Laparoscopic biopsies were collected from suspicious areas on theperitoneum, and curettage samples of the functional layer of theendometrium were carefully collected in single long strips allowingorientation during paraffin embedding. The basal layer of theendometrium was defined as 300μm above the endometrial–myometrialinterface (≤300μm) (Gambino et al., 2002) and the functional layer ofthe endometrium was defined as the area from the epithelial surface tothe basal layer. The basal and functional layers were further dividedinto two sublayers, namely superficial and deeper portions.

The severity of pain was not assessed systematically and prospec-tively in this preliminary observational study, but detailed clinicalinformation was recorded in a standard format. All endometriosispatients were staged according to the revised American Fertility Soci-ety (AFS) score (American Society for Reproductive Medicine, 1996),which ranged from 1 to 1V3. Eighty-six percent of the endometriosissubjects had peritoneal endometriosis alone, while the remainder hadovarian endometriosis with or without endometriosis in other sites.

Menstrual cycle staging was determined on all samples ofendometrium, by a single gynaecological histopathologist (P.R.), andall phases of the cycle were represented: menstrual (days 1–7 of thecycle), proliferative (days 8–14) and secretory (days 15–28) phases ofa normalized 28 day cycle. Of the 35 women with endometriosis, fourwere assessed as being in the menstrual phase, 11 were in the prolifer-ative phase and 20 were in the secretory phase, whereas of the 82women without endometriosis, 12 were assessed as being in the men-strual phase, 36 were in the proliferative phase and 34 were in thesecretory phase of the cycle. None of the women had received medicaltherapy for endometriosis in the 3 months prior to endometrial orhysterectomy sampling.

Immunohistochemistry

After resection, the specimens were immediately fixed in 10% neutralbuffered formalin for ~18–24h, processed and embedded in paraffinwax according to a standard protocol. Each section was cut at 4μmand routinely stained with haematoxylin and eosin (H&E). Antigenretrieval techniques for PGP9.5 and NF were used. Serial sections,also cut at 4μm, were immunostained using an antibody for polyclo-nal rabbit anti-PGP9.5 (dilution 1:700; Dako Cytomation, Sydney,Australia), a highly specific pan-neuronal marker (Lundberg et al.,1988), and monoclonal mouse anti-human NF (dilution 1:200; DakoCytomation), a highly specific marker for myelinated nerve fibres(Schlaepfer, 1987), for 30min at room temperature. Sections werewashed in Tris buffer, incubated with Envision-labelled polymer-APmouse/rabbit for 30min (Dako Cytomation) and stained with perman-ent fast red chromogen (Dako Cytomation) for 10 min. All immunos-taining was carried out on a Dako Autostainer Model S3400 (DakoCytomation Inc., Carpinteria, CA). Images of the sections were cap-tured using an Olympus BX51 digital camera (Olympus, Tokyo,

Japan). We used normal skin as a positive control as it reliablycontains myelinated and unmyelinated nerve fibres expressing bothPGP9.5 and NF. Rabbit immunoglobulin fraction was used as a nega-tive control, the concentration being matched with the concentrationof the PGP9.5 and NF antibodies.

Statistical analysis

The images were captured by using an Olympus BX51 digital camera,and an assessment of nerve fibre density was performed by Image ProPlus Discovery (MediaCybernetics, Silver Springs, MD). Once the fea-tures of the images were controlled, an orthogonal grid mask wassketched above the original images. The sections of the grid were250μm a side. Once the grid was in position, nerve fibres in theendometrium and myometrium, and the total number of squares cover-ing the sections of endometrium and myometrium were counted. Thetotal number of nerve fibres was divided by the total number of squarescovering the endometrium to obtain an average of nerve fibres persquare (each square of 250×250μm). The results were expressed as themean (±SD) number of nerve fibres per mm2 in each specimen from allendometrial and myometrial samples, stained with the polyclonal anti-body against PGP9.5 and monoclonal antibody against NF. Nerve fibredensity between eutopic endometrium from endometriosis patients andnormal women was compared using the Mann–Whitney U-test. Differ-ences were considered to be significant at P<0.05.

Results

Multiple small unmyelinated nerve fibres stained for PGP9.5were present in the functional layer of the endometrium in allendometriosis patients (Figure 1A). Most of the nerve fibreswere in the deeper portion of the functional layer; however,some nerve fibres in the superficial portion were very close(50–180μm) to the endometrial epithelial surface and they oftenaccompanied blood vessels and endometrial glands. In patientswithout endometriosis, no nerve fibres were detected in the func-tional layer of the endometrium (Figure 1B; Table I; P<0.001).

There were also NF-immunoreactive myelinated nervefibres in the deeper portion of the basal endometrium in allendometriosis patients, while in patients without endometriosisno myelinated nerve fibres were detected within the basal layerof the endometrium, with one exception (Table II; P<0.001).This exceptional patient did not have endometriosis, but shehad pelvic pain, menorrhagia, uterine fibroids and chronicinflammation in the cervix. No NF-immunoreactive myeli-nated nerve fibres were detected in the functional layer of theendometrium in patients with or without endometriosis.

There were many more small unmyelinated nerve fibrespresent throughout the basal layer (both deeper and superficial) ofthe endometrium in women with endometriosis (Figure 1C) thanwithout endometriosis. Occasionally, small unmyelinated nervefibres were noted in women without endometriosis but only in thedeeper portion of the basal layer (four women out of 35; Figure1D; Table II; P<0.001). In women with endometriosis, thesenerve fibres often formed obvious thick trunks in the basal layeror at the endometrial–myometrial interface (Figure 1C). Thesetrunks were never seen in women without endometriosis. Thirty-one women without endometriosis showed no nerve fibresthroughout the basal layer of the endometrium. All four womenwithout endometriosis who showed nerve fibres in the deeperportion of the basal endometrial layer had uterine fibroids.



Dow

nloaded from


N.Tokushige et al.

784

Small and large nerve fibres including thick trunks stainedwith PGP9.5 were present in myometrium in women with andwithout endometriosis. There were more nerve fibres stained withPGP9.5 in the myometrium in women with endometriosis than inwomen without (Table III; P<0.001). A small number of nervefibres stained with NF was seen throughout the myometrium ofall women with endometriosis (Table III), but only 16 out of 35women (46%) without endometriosis showed a few nerve fibresstained with NF throughout in the myometrium (P>0.097).

There appeared to be a difference in the patterns of PGP 9.5in the endometrium at different stages of the menstrual cycle.There was a greater density of nerve fibres in the functionaland basal layers of the endometrium in the secretory phase thaneither the menstrual or proliferative phases. The distribution ofnerve fibres was reasonably uniform throughout the functionallayer and basal endometrial layers in women with endometrio-sis- (Figure 1A and C).

There was also a greater density of nerve fibres stained withPGP9.5 in the functional and basal layers of the endometrium thanthe myometrium in women with endometriosis (Tables I–III;P<0.001).

Discussion

This study has demonstrated multiple small unmyelinatednerve fibres in the functional layer of the endometrium in35 women with confirmed endometriosis but in none of82 women without endometriosis. There were also many morenerve fibres in the basal layer of the endometrium in womenwith endometriosis compared with women with no endometri-osis. Some nerve fibres in the basal layer of the endometriumin women with endometriosis were NF immunoreactive andmyelinated, but no NF-immunoreactive nerve fibres werepresent in the functional layer of the endometrium in womenwith or without endometriosis.

Figure 1. Nerve fibres in the functional and basal layers of theendometrium from woman with and without endometriosis. (A)Endometrium from the functional layer of a woman with biopsy-confirmed endometriosis stained for PGP9.5 (magnification×400).Arrows denote multiple small nerve fibres in the functional layer,several within 100μm of the epithelial surface. (B) Endometriumfrom the functional layer of a woman with no visible endometriosisstained for PGP9.5 (magnification×400). No nerve fibres could beidentified. (C)Myometrial–endometrial interface from a woman withbiopsy-confirmed endometriosis stained for PGP9.5 (magnification×200) showing thick nerve fibre trunks stained red. (D) Myometrial–endometrial interface from a woman with no visible endometriosisand with a small nerve fibre in the deeper portion of the basal layerstained with PGP9.5 (magnification×200). (E) Myometrial–endometrialinterface from a woman with biopsy-confirmed endometriosis stainedfor NF (magnification×400) showing several nerve fibres stained red.Scale bars represent 100μm in (A), (B) and (E); and 200μm in (C)and (D).

Table I. Small unmyelinated nerve fibre density in the functional layer of endometrium (per mm2, mean±SD) in women with and without endometriosis

Nerve fibres were stained with PGP9.5.P<0.001 for endometriosis versus no endometriosis.

n Mean±SD Range

Hysterectomy specimens: endometriosisMenstrual 1 7 7Proliferative 2 6 5–6Secretory 7 15±5 9–23Total 10 11±5Hysterectomy specimens: no endometriosisMenstrual 5 0 0Proliferative 16 0 0Secretory 14 0 0Total 35 0Curettage specimens: endometriosisMenstrual 3 7±1 6–8Proliferative 9 5±1 3–6Secretory 13 13±6 7–30Total 25 10±7Curettage specimens: no endometriosisMenstrual 7 0 0Proliferative 20 0 0Secretory 20 0 0Total 47 0 0



Dow

nloaded from



785

It has been known for some time that the myometrium, theendometrial–myometrial interface and the deeper portion of thebasal endometrium can be innervated by nerve fibres, but thatnerve fibres are absent from the superficial two-thirds of theendometrium (the functional layer) in the normal human uterus(Coupland, 1969). The function of these nerve fibres in normalbasal endometrium is not well understood. However, some ace-tylcholinesterase (AChE)-immunoreactive nerve fibres weredetected in the basal layer of normal human endometrium(Coupland, 1969). Neuropeptide - (NPY), substance P- (SP),vasoactive intestinal peptide- (VIP) and neurotensin (NT)-immunoreactive nerve fibres were also present in normal humanendometrium (Heinrich et al., 1986). Yet little is known aboutthe functions of these nerve fibres in human endometrium.

Some researchers have studied nerve fibres in normalendometrium in animals. Calcitonin gene-related peptide-(CGRP) (Shew et al., 1990), AChE- and NPY-immunoreactivenerve fibres were detected in rat endometrium (Papka et al., 1985)and many CGRP-immunoreactive nerve fibres were near the sur-face epithelium (Shew et al., 1990). SP and CGRP were alsolocalized in myometrial nerves in a rat model (Shew et al., 1990).

A variety of clinical conditions including endometriosis, pelvicinflammatory disease, uterine fibroids, adhesions and uterinecontractions can cause pelvic pain. Various investigators havedemonstrated nerve fibres in endometriotic plaques, abdominaladhesions and myometrium (Tulandi et al., 1998; Anaf et al.,2000; Sulaiman et al., 2001; Tamburro et al., 2003; Berkleyet al., 2004). Endometriotic plaques in a rat model were inner-vated by SP, CGRP (sensory C and Aδ fibres) and vesicularmonoamine transporter (VMAT) fibres (Berkley et al., 2004).

Nerve fibres were present in human peritoneal endometrioticplaques, and transforming growth factor (TGF-β1) wasexpressed in the nerve fibres only in red and white lesions(Tamburro et al., 2003). Rectovaginal endometriotic nodulesfrom women with severe dysmenorrhoea and deep dyspareuniashowed the presence of nerve fibres staining with the non-specific antibody, S-100 protein (Anaf et al., 2000). Humanperitoneal adhesions contained synaptophysin- CGRP-, SP-,vasoactive intestinal peptide- (VIP)- and tyrosine hydroxylase(TH)-immunoreactive nerve fibres (Sulaiman et al., 2001) and78% of endometriosis-related adhesions showed NF-immuno-reactive nerve fibres (Tulandi et al., 1998).

SP- and CGRP-immunoreactive nerve fibres were present inhuman myometrium (Samuelson et al., 1985) and SP-inducedcontractions were inhibited by CGRP. It has been reported thatwomen with endometriosis had uterine contractions withhigher frequency, amplitude and basal pressure tone duringmenses (Bulletti et al., 2002); therefore, SP- and CGRP-immu-noreactive nerve fibres may play a role in the pain of womenwith endometriosis.

The pathogenesis of endometriosis is not well understood,but there are numerous theories, including implantation ofendometrial fragments transported via the Fallopian tubes byretrograde menstruation (Sampson, 1927, 1940), mechanicaltransplantation (Allen et al., 1954; Levander and Normann,1955), lymphatic and vascular metastasis (Halban, 1924;Jubanyik and Comite, 1997), direct extension or invasion (Cullen,1908), coelomic metaplasia (Ridley, 1968, Vasquez, 1984),embryonic rests (Von Recklinghausen, 1896; Russell, 1899)and a composite theory (Javert, 1949; Nisolle and Donnez,

Table II. Small unmyelinated and large myelinated nerve fibre densities in the basal layer of endometrium (per mm2, mean±SD) in women with and without endometriosis

Unmyelinated and myelinated nerve fibres were stained with PGP9.5; myelinated nerve fibres were stained with NF.P<0.001 for endometriosis versus no endometriosis.

Endometriosis No endometriosis

n PGP9.5 NF n PGP9.5 NF

Mean±SD Range Mean±SD Range Mean±SD Range Mean±SD Range

Menstrual 1 12 12 1 1 5 0 0–1 0 0–1Proliferative 2 11 9–18 2 1–2 16 0 0–1 0 0Secretory 7 21±8 11–80 3±2 1–7 14 0 0–1 0 0Total 10 18±8 2±2 35 0 0

Table III. Small unmyelinated and large myelinated nerve fibre densities in myometrium (per mm2, mean±SD) in women with and without endometriosis

Unmyelinated and myelinated nerve fibres were stained with PGP9.5; myelinated nerve fibres were stained with NF.P<0.001 for PGP9.5; P>0.097 for NF.

Endometriosis No endometriosis

n PGP9.5 NF n PGP9.5 NF

Mean±SD Range Mean±SD Range Mean±SD Range Mean±SD Range

Menstrual 1 2.28 2.28 0.02 0.02 5 0.96±0.55 0.29–1.76 0.03±0.02 0.01–0.06Proliferative 2 2.62 2.51–2.72 0.03 0.03 16 0.49±0.34 0.15–1.20 0.01±0.01 0.01–0.03Secretory 7 3.58±1.36 2.21–6.24 1.00±0.03 0.96–1.03 14 1.12±1.06 0.17–3.13 0.03±0.02 0.01–0.09Total 10 3.26±1.23 0.71±0.47 35 0.89±0.78 0.02±0.02



Dow

nloaded from


N.Tokushige et al.

786

1997). Most researchers support variations of the implantationtheory, which postulates that peritoneal endometriotic lesionsresult from the reflux of viable endometrial tissue fragments orcells through the Fallopian tubes which then implant and growon the peritoneal surface and pelvic organs.

If the implantation theory is true, the nerve fibres inendometriotic plaques may originate from nerve fibre progeni-tors in the functional layer of the endometrium or, more prob-ably, from ingrowth of local nerve fibres due to the secretion ofnerve growth factors (NGFs) and increased expression of twoNGF receptors, namely Trk-A and p75, by the implantingendometrium. It is not yet clear what neural stimuli may initi-ate the severe pain sensations which some women experiencewith endometriosis. However, a number of molecules that arecapable of stimulating nerve fibres, such as tumour necrosisfactor-α (TNF-α) (Bergqvist et al., 2001), prostaglandin I2(Ylikorkala and Viinikka, 1983), prostaglandin E2 and F2α (DeLeon et al., 1988), and NGF (Anaf et al., 2002) can be releasedfrom endometriotic plaques. It seems highly probable that theendometrium of women destined to develop endometriosis isfunctionally abnormal (Lessey et al., 1994; Ota and Tanaka,1997; Healy et al., 1998). It may also be programmed tosecrete large amounts of nerve trophic factors, resulting in theremarkable ingrowth of unmyelinated small nerve fibres dem-onstrated in this study. This may also be translated intoingrowth of nerve fibres into new endometriotic plaques on theperitoneum or ovary. These nerve fibres may have an import-ant role in the mediation of pain symptoms. No other satisfac-tory theory has yet been advanced for the genesis of painsymptoms in women with endometriosis.

This demonstration of small nerve fibres in the functionallayer of eutopic endometrium of women with endometriosis isso striking in the present study that we believe it could becomea relatively simple surrogate marker of this condition usingendometrial biopsies. It is even possible that an endometrialbiopsy may demonstrate nerve fibres in women destined todevelop the condition later in life. Much further research isrequired to define possible roles and mechanisms of thesenerve fibres in endometriosis, as well as defining the presenceof uterine nerve fibres in other gynaecological conditions.

AcknowledgementsThe authors thank Lawrence Young (Dako Cytomation,Australia) for technical support. This study was supported byresearch funding from the Department of Obstetrics andGynaecology, University of Sydney.

References

Al-Fozan H, Bakare S, Chen MF and Tulandi T (2004) Nerve fibres in ovariandermoid cysts and endometriomas. Fertil Steril 82,230–231.

Allen E, Perterson LF and Campbell ZB (1954) Clinical and experimentalendometriosis. Am J Obstet Gynecol 68,356–375.

American Society for Reproductive Medicine (1996) Revised American Society forReproductive Medicine classification of endometriosis. Fertil Steril 67,817–821.

Anaf V, Simon P, El Nakadi I, Fayt I, Buxant F, Simonart T, Peny MO andNoel JC (2000) Relationship between endometriotic foci and nerves in rec-tovaginal endometriotic nodules. Hum Reprod 15,1744–1750.

Anaf V, Simon P, Nakadi IEI, Fayt I, Simonart T, Buxant F and Noel JC(2002) Hyperalgesia, nerve infiltration and nerve growth factor expression

in deep adenomyotic nodules, peritoneal and ovarian endometriosis. HumReprod 17,1895–1900.

Bergqvist A, Bruse C, Carlberg M and Carlstrom K (2001) Interleukin 1beta,interleukin-6, and tumour necrosis factor-alpha in endometriotic tissue andin endometrium. Fertil Steril 75,489–495.

Berkley KJ, Dmitrieva N, CurtisK and Papka RE (2004) Innervation of ectopicendometrium in a rat model of endometriosis. Proc Natl Acad Sci USA101,11094–11098.

Binkovitz LA, King BF and Ehman RL (1991) Sciatic endometriosis: MRappearance. J Comput Assist Tomogr 15,508–510.

Bulletti C, De Ziegler D, Polli V, Del Ferro E, Palini S and Flamigni C (2002)Characteristics of uterine contractility during menses in women with mild tomoderate endometriosis. Fertil Steril 77,1156–1161.

Coupland RE (1969) The distribution of cholinergic and other nerve fibres inthe human uterus. Postgrad Med J 45,78–79.

Cullen TS (1908) Adenomyoma of the Uterus. WB Saunders, Philadelphia,pp. 222–227.

De Leon FD, Vijayakumar R, Rao CV and Yussman M (1988) ProstaglandinF2 alpha and E2 release by peritoneum with and without endometriosis. Int JFertil 33,48–51.

Gambino LS, Wreford NG., Bertram JF, Dockery P, Lederman F and RogersPA (2002) Angiogenesis occurs by vessel elongation in proliferative phasehuman endometrium. Hum Reprod 17,1199–1206.

Halban J (1924) Metastatic hysteroadenosis. Wien Klin Wochenschr 37, 1205–1206.Healy DL, Rogers PA, Hii L and Wingfield M (1998) Angiogenesis: a new

theory for endometriosis. Hum Reprod Update 4,736–740.Heinrich D, Reinecke M and Forssmann WG (1986) Peptidergic innervation of

the human and guinea pig uterus. Arch Gynecol 237,213–219.Javert CT (1949) Pathogenesis of endometriosis based on endometrial homeo-

plasia, direct extension, exfoliation and implantation, lymphatic and hema-togenous metastasis. Including five case reports of endometrial tissue inpelvic lymph nodes. Cancer 2,399–410.

Jubanyik KJ and Comite F (1997) Extrapelvic endometriosis. Obstet GynecolClin North Am 24,411–440.

Lessey BA, Castelbaum AJ, Sawin SW, Buck CA, Schinnar R, Bilker W andStrom BL (1994) Aberrant integrin expression in the endometrium ofwomen with endometriosis. J Clin Endocrinol Metab 79,643–649.

Levander G and Normann P (1955) The pathogenesis of endometriosis. Anexperimental study. Acta Obstet Gynecol Scand 34,366–398.

Lundberg LM, Alm P, Wharton J, Polak JM (1998) Protein gene product 9.5(PGP9.5). A new neuronal marker visualizing the whole uterine innervationand pregnancy-induced and developmental changes in the guinea pig. Histo-chemistry 90,9–17.

Nisolle M and Donnez J (1997) Peritoneal endometriosis, ovarian endometrio-sis, and adenomyotic nodules of the rectovaginal septum are three differententities. Fertil Steril 68,585–596.

Ota H and Tanaka T (1997) Integrin adhesion molecules in the endometrialglandular epithelium in patients with endometriosis or adenomyosis.J Obstet Gynaecol Res 23,485–491.

Papka RE, Cotton JP and Traurig HH (1985) Comparative distribution of neu-ropeptide tyrosine-, vasoactive intestinal polypeptide-, SP-immunoreactive,acetylcholinesterase-positive and noradrenergic nerves in the reproductivetract of the female rat. Cell Tissue Res 242,475–490.

Ridley JH (1968) The histogenesis of endometriosis: a review of facts and fan-cies. Obstet Gynecol Surv 23,1–35.

Russell WW (1899) Aberrant portions of the müllerian duct found in an ovary.Ovarian cysts of müllerian origin. Bull Johns Hopkins Hospital 10,8–10.

Sampson JA (1927) Peritoneal endometriosis due to the menstrual dissemination ofendometrial tissue into the peritoneal cavity. Am J Obstet Gynecol 14,422–469.

Sampson JA (1940) The development of the implantation theory for the originof peritoneal endometriosis. Am J Obstet Gynecol 40,549–557.

Samuelson UE, Dalsgaard CJ, Lundberg JM and Hokfelt T (1985) Calcitoningene-related peptide inhibits spontaneous contractions in human uterus andfallopian tube. Neurosci Lett 62,225–230.

Schlaepfer WW (1987) Neurofilaments: structure, metabolism and implica-tions in disease.J Neuropathol Exp Neurol46,117–129.

Shew RL, Papka RE and McNeill DL (1990) Calcitonin gene-related peptidein the rat uterus: presence in nerves and effects on uterine contraction.Peptides 11,583–589.

Sulaiman H, Gabella G, Davis MSc C, Mutsaers SE, Boulos P, Laurent GJ andHerrick SE (2001) Presence and distribution of sensory nerve fibers inhuman peritoneal adhesions. Ann Surg 234,256–261.

Tamburro S, Canis M, Albuisson E, Dechelotte P, Darcha C and Mage G(2003) Expression of transforming growth factor beta1 in nerve fibres is



Dow

nloaded from



787

related to dysmenorrhoea and laparoscopic appearance of endometrioticimplants. Fertil Steril 80,1131–1136.

Tulandi T, Chen MF, Al-Took S and Watkin K (1998) A study of nerve fibersand histopathology of postsurgical, postinfectious, and endometriosis-related adhesions. Obstet Gynecol 92,766–768.

Vasquez G, Cornillie F, Brosens IA (1984) Peritoneal endometriosis: scanningelectron microscopy and histology of minimal pelvic endometriotic lesions.Fertil Steril 42,696–703.

Von Recklinghausen F (1896) Adenomyomas and cystadenomas of the wall ofthe uterus and tube: their origin as remnants of the wolffian body. Wien KlinWochenschr 8,530.

Ylikorkala O and Viinikka L (1983) Prostaglandins and endometriosis. ActaObstet Gynecol Scand Suppl113,105–107.

Submitted on August 10, 2005; resubmitted on September 28, 2005; acceptedon September 29, 2005



Dow

nloaded from


hum. reprod.-2006-tokushige-782-7

Documents