{"paper_id":"5f98cd89-5da3-4aad-9659-795a8cdc8a88","body_text":"ORIGINAL ARTICLE\nMarc Possover Æ Kerstin Rhiem Æ Vito Chiantera\nThe ‘‘neurologic hypothesis’’: a new concept in the pathogenesis\nof the endometriosis?\nReceived: 19 August 2004 / Accepted: 29 November 2004 / Published online: 13 May 2005\n/C211Springer-Verlag Berlin / Heidelberg 2005\nAbstract To cartograph the retroperitoneal infiltration\nof deep-infiltrating endometriosis of the rectovaginal\nspace, we report on 467 patients who underwent lapa-\nroscopic/vaginal surgery for deep-infiltrating adenomy-\nosis of the rectovaginal space. Exact localisation of the\nlocoregional extension and of secondary infiltrating lo-\ncalisation where noted. The cervix and the rectovaginal\nligament were mostly involved, while isolated lesions of\nthe rectovaginal space were very rare. Comparisons of\nthe most involved sites show an absolute correlation\nwith the anatomical repartition of the pelvic sympathetic\nnervous system. We postulate a new ‘‘neurologic the-\nory’’ which could be one more explanation for the\ndevelopment of adenomyosis of the rectovaginal space\nand maybe the peritoneum.\nKeywords Endometriosis Æ Laparoscopic-vaginal\nsurgery Æ Sympathetic nerve system\nIntroduction\nAdenomyosis behaves retroperitoneally like a tumor and\ncan affect various retroperitoneal structures. It remains\nunclear why and with which mechanisms this endome-\ntriosis infiltrates the retroperitoneum. The aim of our\nwork was to establish an invasion pattern of the retro-\nperitoneum.\nMaterial and methods\nWe present a study of all of our patients on whom we\nperformed radical surgery for adenomyosis of the rect-\novaginal septum since 1997. All of the surgical proce-\ndures began with a rectovaginal examination of the\npatient. Particular attention was paid to the estimation\nof involvement/stenosis of the rectum and involvement\nof the pelvic wall during rectal palpation. When the\nvagina was involved, the nodule was first dissected on\nthe vaginal route and followed by laparoscopy. When\nthe vagina was determined free of disease during pal-\npation, the procedure began directly with a laparoscopy.\nA minute inspection of the entire abdominal cavity and\nof the pelvis was performed. When there was retroperi-\ntoneal infiltration or rectosigmoid infiltration, the ret-\nroperitoneal route was primarily chosen, and the exact\ndepth and side/area of infiltration was documented. Our\ntechnique of deep anterior rectum resection/anastomosis\nhas been previously described [ 1]. Directly after the\nprocedure, infiltration and the side of the infiltration of\nthe following sites were documented:\n– Cervix\n– Vagina\n– Pouch of Douglas\n– Inferior hypogastric plexus right/left (rectovaginal\nligament)\n– Ilio- and ischio-coccegeus muscles right/left\n– Sacral plexus and/or sciatic nerve right/left\n– Vascular part of the cardinal ligament right/left (from\nthe uterine arteria to the medial rectal artery)\n– Neural part of the cardinal ligament right/left (Nervi\nsplanchnici pelvini)\n– Ureter right/left\nNo financial support was received for this study.\nM. Possover ( &) Æ K. Rhiem\nDepartment of Obstetrics and Gynecology, University of Cologne,\nKerperner Street 34, 50924 Cologne, Germany\nE-mail: MarcPossover@aol.com\nTel.: +49-221-4677-1301\nFax: +49-221-4677-1309\nV. Chiantera\nUniversity of Naples II, Naples, Italy\nPresent address : M. Possover\nDepartment of Gynecology,\nElisabeth-Krankenhaus Ko¨ln-Hohenlind, Werthmannstrasse 1,\n50935 Cologne, Germany\nGynecol Surg (2005) 2: 107–111\nDOI 10.1007/s10397-004-0070-8\n\n– Rectum (mesenterial/antimesenterial)\n– Sigmoid (mesenterial/antimesenterial)\n– Ileo-coecum/appendix\n– Bladder-wall\n– Others\nAll these data were registered prospectively using\nExcel (Microsoft, Redmond, WA, USA).\nResults\nSince December 1997 we have performed radical lapa-\nroscopic or laparoscopic-assisted vaginal surgery for\ndeep-infiltrating endometriosis of the rectovaginal space\non 467 consecutive patients. All surgical procedures\nwere performed by the author himself, originally at the\nDepartment of Gynecology at the University of Jena\nand since 2001 at the Women’s Hospital at the Univer-\nsity of Cologne. All procedures were performed lapa-\nroscopically except for one patient with extended\ninfiltration of the anterior abdominal wall where a lap-\narotomy was required. Table 1 shows the intraoperative\ndata and the different surgical procedures.\nIsolated infiltration of the rectovaginal space without\ninfiltration of the Douglas pouch and/or the rectova-\nginal ligaments and/or the cervix was found in only three\npatients (0.6%). In all other patients (99.4%), infiltra-\ntion of the posterior cervix and at least one rectovaginal\nligament at its insertion at the cervix was found. We\nnever observed a single lesion of the rectovaginal space\nwith infiltration of the Douglas pouch without involve-\nment of at least one of the rectovaginal ligaments. When\nthe rectovaginal ligaments were infiltrated, both sides\nwere involved in 266 patients (57%), only on the right\nside in 70 patients (15%) and only on the left side in 128\npatients (27.4%).\nUreteral involvement/stenosis was found in 29 pa-\ntients. In one patient this ureteral lesion was found\nisolated at the level of its crossing with the left common\niliac artery, while in all other patients the ureter was\ninvolved in a conglomerate lesion at the level of the cross\nbetween the ureter itself and the uterine artery. The\nureteral involvement was found on the left side in 27\npatients and on both sides in one patient. One further\npatient presented with an isolated ureteral stenosis on\nthe right side, but in her anamnesis, ureteral resection/\nanastomosis on the left side was relevant.\nExtension to the neural part of the cardinal ligament\n(Nervi splanchnici pelvini ) was found in 31 patients, 17\non the left side and 14 on the right side. No bilateral\ninfiltration was observed at this level.\nResection of a part of the iliococcygeus and/or is-\nchiococcygeus muscles was required on 32 patients (11\non the left and 21 on the right). Infiltration of the sciatic\nnerve itself was found on the right side in five patients\nand on the left side in two further patients.\nDeep and extended infiltration of the rectum with\nrectal stenosis was found in 151 patients. In 17 of these\npatients, associated infiltration of the sigmoid was ob-\nserved, while in three further patients, isolated stenosis\nof the sigmoid without rectal infiltration was found.\nNo isolated mesial infiltration of the bowel was found.\nIn three patients with rectum infiltration, an associated\nextended infiltration of the ileocoecum was noted and\nrequired a double intestinal resection. Isolated involve-\nment of the appendix was only found in four patients.\nIn patients where an anterior rectum resection/anas-\ntomosis was performed, no recurrence in the rectova-\nginal space, in both rectovaginal ligaments or in the\nvagina has been observed to date. Where an isolated\nresection of a nodule of the rectovaginal space was\nperformed but without enlarged resection of both rect-\novaginal ligaments, 33 patients presented with a recur-\nrence at the level of the rectovaginal ligament or of the\nupper part of the vagina.\nDiscussion\nWhen we consider the frequency of infiltration in the\ndifferent sites, the cervix was nearly always involved,\nwhile a single lesion isolated to the rectovaginal space\nwas very rare. It seems that the adenomyosis of the\nrectovaginal space is due more to an invasion coming\nfrom the cervix—like a ‘‘benign cancer’’ of the cer-\nvix—than a coelomic metaplasia directly in the rect-\novaginal space [ 2]. The endometriosis may grow and\nprogress through direct extension and first infiltrates\nalong the rectovaginal ligament. From there the disease\nis able to leave the rectovaginal ligament—statistically\nmore frequently on the left side—and extend:\n– Medially to the rectovaginal space and from there\ncranially to the Douglas pouch, ventrally to the va-\ngina and dorsally to the rectum.\n– Laterally/cranially along the uterine artery to the pars\nvasculosa of the cardinal ligament and consequently\nto the ureter (mostly on the left side)\n– Laterally/caudally to the pelvic wall and the sacral\nplexus\nTable 1 Operative data ( n=467)\nData\nMean age (years) 26.3\nNumber of previous surgeries (average) 6.5\nOperative time (average)\nWith bowel resection 169.4 min\nWithout bowel resection 125.3 min\nBloodtransfusion ( n) n=4\nFurther laparoscopic procedures\nBladder-resection n=22\nUreter-resection/anastomosis n=2\nUreter-resection/ureterocystoneostomy n=15\nRectum-resection/anastomosis n=151\nSigmoide-resection/anastomosis n=3\nHemicolectomy/anastomosis n=1\nCoecum-resection/ileoascendostomy n=3\nHysterectomy n=6\n108\n\n– Dorsally to the rectum at its lateroanterior face\nWhen we compare the cartography of the most fre-\nquent infiltration sites of the retroperitoneal adenomy-\nosis with the topographic anatomy of the pelvic\nsympathetic nerves, we discover that both are identical:\n– The cervix: the distribution of autonomous nerves in\nthe genital organs is more extensive in the proximal\nvagina and the uterine cervical myometrium than in\nthe uterus itself and is quite exclusively of a sympa-\nthetic type [ 3–5].\n– The rectovaginal ligament: the s.c. sacrouterine liga-\nments are in fact not ligaments but the inferior\nhypogastric plexus itself. At this level—cervix and\nupper third of the vagina—this plexus contains sym-\npathetic fibres exclusively.\n– The upper part of the vagina: most of the nerves that\njoin the vagina from the inferior hypogastric plexus\nenter its dorsal and middle parts. No nerves of mac-\nroscopic size can be traced from this plexus to the\ncaudal part of the vagina [ 6].\n– The trigonom of the bladder: the invasion coming\nfrom the anterior part of the cervix mostly involves\nthe upper part of the trigonom where the sympathetic\nnerves enter the bladder.\n– Parametry along the uterine artery: from the plexus\npelvicus, sympathetic nerves are expanded laterally\nalong the uterine artery to the terminal ureter [ 6] and\nto the lateral peritoneum of the ovarian fossa\n– The ventrolateral portion of the rectum: sympathetic\nnerves coming from the pelvic plexus enter the rectum\nat its’ anterolateral part.\nThus it seems that a correlation exists between the\nanatomical site of the pelvic sympathetic nerves and\nthe infiltrating endometriosis. There could be two dif-\nferent hypotheses to explain this correlation. The first\ncould be a neurotropism of the endometriosis for the\nsympathetic nerves. In this theory, the endometriosis\nshould expand along the pelvic nerves, and the par-\nticular localisation of endometriosis on the cervix, the\nupper vagina and the rectovaginal ligaments could be\ndue to the higher concentration of sympathetic nerves\nin these anatomical structures.\nThe second explanation could be that the sympathetic\nnerves play a direct role in the local development of the\nendometriosis. In reviewing the literature of the sympa-\nthetic nervous system, we found that the entire peritoneal\npelvic cavity as well as the pelvic organs including the\nuterine vasculature are innervated by several subpopu-\nlations of sympathetic nerves [ 7, 8] but there are two\nparticularities for the pelvic sympathetic innervation:\n1. The first particularity is that in contrast to other parts\nof the peripheral nervous system, the uterine adren-\nergic innervation is highly influenced by endocrine\nfactors (e.g. sex steroids) [ 9]. It was shown that the\nlevel of norepinephrine transmitter varies during the\noestrous cycle [ 10]. Even more pronounced are the\nchanges seen during the course of pregnancy, when\nthe adrenergic nerves in the fetus-containing uterine\nhorn undergo structural degeneration which is almost\ncomplete at term pregnancy; during the puerperal\nperiod, this adrenergic transmitter synthesis increases\nvery slowly over the following months [ 11]. Thus, the\nuterus is able to alter its state of innervation through\na hormonally controlled variation in the production\nof such neurotropic material essential for the main-\ntenance or growth of a normal innervation. This\nphenomenon is of great importance because it is well\nknown that endometriosis is decreasing or disap-\npearing during the pregnancy while the level of oes-\ntrogen is increasing: It may be hypothesized that this\ndecreasing of the endometriotic activity during the\npregnancy and the puerperal period is directly the\nconsequence not of the hormonal situation but of the\ndisapparition of this uterine adrenergic transmitter\nsynthesis. This is further in agreement with the\nobservation that ovariectomy, which reduces the\nweight of the cervix and the uterine horns, also re-\nduces total organ adrenergic transmitter synthesis\n[12]. On these grounds it may be further hypothesised\nthat if this reduction of adrenergic activity should be\nable to reduce the local endometriotic activity, in\ncontroversy the increasing of this activity could be\nresponsible for the development and/or growth of the\nendometriosis. Such a phenomenon is well known\nfrom the adaptation doctrine of Selye who has shown\nthat the sympathetic activity could increase under\nexternal influence like that of the stress and could\nplay a real role in psychophysiological aspects of\ntumor development [ 13]. Why the development of the\nendometriosis is mostly show on the cervix could be\nexplained by two histologic particularities of this\nanatomical region: the first is that a very rich network\nof sympathetic adrenergic nerves innervate blood\nvessels as well as the cervical myometrium immedi-\nately beneath the epithelium. The second particular-\nity is that during the mens, at the junction between\nthe uterine endocrine and the cervical squamous cell\nepithel, superficial epithel cells are freed from their\nbasal attachment. This is of importance because these\nendocrine cells are still alive and have the ability to\ninfiltrate and proliferate deep in the cervix. It is our\nsecond hypothesis that changes in the synthesis, re-\nlease, or proliferation of uterine adrenergic trans-\nmitter induced by humoral and/or local factors could\ninfluence the proliferation of these cells in the cervix\nending in the formation of endometriosis.\n2. The second particularity is that two of the subtypes of\nthe pelvic sympathetic nerves—the s.c. neuropeptide-\nY autonomous (NPY) nerves and the vasoactive\nintestinal polypeptide (VIP) nerves—show a particu-\nlar localisation for the cervix, the rectovaginal liga-\nment and the upper vagina—where we have mostly\nfound the infiltrating endometriosis [ 14–16]. This is of\nimportance because these autonomic nerves, espe-\n109\n\ncially the NPY nerves, are one of the most powerful\nangiogenic factors in the body and are involved in the\nneoangiogenesis of the pelvic organ: they are able to\ncause vascular smooth muscle cell proliferation and\nthe stimulation of endothelial cell adhesion to matrix,\nmigration, proliferation, capillary tube formation on\nmatrigel and aortic sprouting [ 17]. Thus the angio-\ngenic effect of the NPY and VIP nerves could be di-\nrectly involved in the growth of endometriosis as\nangiogenic factors are able to establish, grow and\ninvade endometriotic implants [ 18] especially in the\npronounced stage IV [ 19–21].\nThus it is our theory that the sympathetic nervous\nsystem is directly involved in the establishment and\ngrowth of the endometriosis.\nAs a consequence we postulate that interference with\nthe local pelvic sympathetic innervation—as in the\npregnancy—may constitute novel therapeutic opportu-\nnities for the prevention, amelioration or treatment of\npelvic endometriosis. In comparison to antiangiogenic\ncompounds alone which are being tested as therapeutic\nagents in the treatment of endometriosis [ 22, 23], control\nof the local sympathetic nervous system should present\none more approach: the direct control of the sensation of\npain and dyspareunia by stopping the transport of sen-\nsitive information to the spinal cord and consequently to\nthe central nervous system [ 24].\nDifferent methods of control on the local sympathetic\nnervous system should be possible, one of which is the\nsurgical transection of these nerves. In our study, radical\nsurgery on extended lesions with bowel resec-\ntion—where normally the risk of recurrence appears to\nbe higher than in small localised lesions—showed better\nresults in terms of recurrence than in localised resection\nof smaller lesions: in both situations we had excised the\nlesion with a histologically proven free margin but in\nsituations where there was extended endometriosis of the\nrectovaginal space with rectum infiltration, resection of\nboth rectovaginal ligaments and the upper part of the\ndorsal vaginal fornix was systematically required and\nperformed while in smaller lesions only a reducing\nresection of the inferior hypogastric plexus was required.\nThus in this radical surgery, the lowest rate of local\nrecurrence could be explained by the higher bilateral\ndestruction of the local vascularisation and of the cer-\nvicovaginal autonomous sympathetic innervation (infe-\nrior hypogastric plexus).\nThis neural theory could be extended to other locali-\nsations of endometriosis since this adrenergic pelvic\ninnervation is also derived from the prevertebral and\nparavertebral ganglia via the hypogastric nerves—sig-\nmoidal localisation for endometriotic implants—as well\nas via nerves running in the suspensory ligament to the\ntubal ends of the uterine horns—ovarial, tubal and peri-\ntoneal endometriosis of the fossa ovarica [ 25].\nOur ‘‘neural theory’’ and its therapeutical implications\nhave to be confirmed by fundamental studies in the\nlaboratory.\nReferences\n1. Possover M, Diebolder H et al (2000) Laparoscopically assisted\nvaginal resection of rectovaginal endometriosis. Obstet Gyne-\ncol 96:304–307\n2. Craig A (1999) Current concepts in the pathogenesis of endo-\nmetriosis. Clin Obstet Gynecol 42:566–585\n3. Adam N, Schenk EA (1979) Autonomic innervation of the rat\nvagina, cervix and uterus and its cyclic variation. Am J Obstet\nGynec 104:508–516\n4. Garfield RE (1986) Structural studies of innervation on non-\npregnant rat uterus. Am J Physiol 251:C41–C54\n5. Sato S, Hayashi RH et al (1989) Mechanical responses of the\nrat uterus, cervix and bladder to stimulation of hypogastric and\npelvic nerves in vivo. Biol Reprod 40:209–219\n6. Kuntz A (1953) Nervous system. Lea and Febiger, Philadel-\nphia, pp 276–303\n7. Fallgren B, Ekblad E et al (1989) Co-existence of neuropeptides\nand differential inhibition of vasodilator responses by neuro-\npeptide-Y in guinea pig uterine arteries. Neurosci Lett 100:71–\n76\n8. Bell C (1968) Dual vasoconstrictor and vasodilator innervation\nof the uterine arterial supply in the guinea pig. Circulatory Res\n23:279–289\n9. Owman CH, Sjo¨berg NO (1977) Influence of pregnancy and sex\nhormones on the system of short adrenergic neurons in the\nfemale reproductive tract. Experta Medica Int Congr Ser\n402:205–209\n10. Thobert G, Alm P et al (1977) Regional distribution of auto-\nnomic nerves in guinea-pig uterus. Am J Physiol 233:C25–C34\n11. Owman Ch, Alm P, et al (1975) Variations in the level of\nuterine norepenephrine during pregnancy in the guinea-pig. Am\nJ Obstet Gynecol 122:961–964\n12. Alm P, Bjo ¨rlund A et al (1979) Tyrosine hydroxylase and\nDOPA decarboxylase activities in the guinea-pig uterus: further\nevidence for functional adrenergic denervation in association\nwith pregnancy. Neuroscience 4:145–154\n13. Weinstocks C (1984) Psychophysiological aspects of cancer.\nMed Hypotheses 15:369–383\n14. Sternquist M, Emson P et al (1983) Neuropeptide-Y in the\nfemale reproductive tract of the rat: distribution of nerve fibres\nand motor effects. Neurosci Lett 39:279–284\n15. Tenmoku S, Ottesen B et al (1988) Interaction of NPY and VIP\nin regulation of myometrial blood flow and mechanical activity.\nPeptides 9:269–275\n16. Alm P, Alumets J et al (1980) Origin and distribution of VIP\n(vasoactive intestinal polypeptide) nerves in the genito-urinary\ntract. Cell Tissue Res 205:337–347\n17. Zukowska Z, Grant DS et al (2002) L. Neuropeptide y: a novel\nmechanism for ischemic angiogenesis. Trends Cardiovasc Med\n13:86–92\n18. Taylor RN, Lebovic DI et al (2002) Angiogenic factors in\nendometriosis. Ann N Y Acad Sci 955:89–100\n19. Barcz E, Rozewska ES et al (2002) Angiogenic activity and IL-8\nconcentrations in peritoneal fluid and sera in endometriosis. Int\nJ Gynaecol Obstet 3:229–235\n20. Han Y, Zhou Y et al (2002) Study on the expression of vascular\nendothelial growth factor in patients with adenomyosis of the\nuterus. Zhonghua Fu Chan Ke Za Zhi 37:539–541\n21. Ria R, Loverro G et al (2002) Angiogenesis extent and\nexpression of matrix metalloproteinase /C02 and /C09 agree with\nprogression of ovarian endometriomas. Eur J Clin Invest\n32:199–206\n22. Reynolds LP, Grazul-Bilska AT et al (2002) Angiogenesis in\nthe female reproductive organs: pathological implications. Int\nJ Exp Pathol 83:151–163\n23. Dabrosin C, Gyorffy S et al (2002) Therapeutic effect of\nangiostatin gene transfer in a murine model of endometriosis.\nAm J Pathol 161:909–918\n110\n\n24. Traurig HH, Mayerhofer A et al (1992) Synaptophysin and\nchromogranin-A are expressed in fetal and adult female rat\nparacervical ganglion autonomic neurons and SIF cells. Soc\nNeurosci Abstr 18:52\n25. Thorbert G, Alm P et al (1978) Cyclic and steroid-induced\nchanges in adrenergic neurotransmitter level of guinea-pig\nuterus. Acta Obstet Gynecol Scand 57:45–48\n111","source_license":"CC0","license_restricted":false}