Mediastinal Cyst of Müllerian Origin: Evidence for Developmental Endosalpingiosis

Müllerianosis, the triad of endometriosis, endosalpingiosis and endocervicosis, has long been recognised but still is a challenging subject regarding its histogenesis. Several theories have been developed to explain the pathogenesis of endometriosis: there are those proposing that uterine endometrium is the site of origin, which either by retrograde menstruation or by lymphatic/haematogenous spread reaches distant sites; and those proposing that endometriotic foci arise from tissue other than uterine, such as müllerian rests, metaplasia of the mesothelium or from bone marrow progenitor cells.1Burney R.O. Giudice L.C. Pathogenesis and pathophysiology of endometriosis.Fertil Steril. 2012; 98: 511-519Abstract Full Text Full Text PDF PubMed Scopus (886) Google Scholar Endocervicosis, the presence of endocervical type glands into the bladder, colon or lymph nodes, probably derives from embryological rests, but when encountered deep in the uterine cervix, is mostly related to caesarian section as a form of diverticulosis (isthmocele).2Mobarki M. Karpathiou G. Forest F. Corsini T. Peoc'h M. Endocervicosis of the uterine cervix.Int J Gynecol Pathol. 2016; 35: 475-477Crossref Scopus (7) Google Scholar Finally, endosalpingiosis, the finding of glands lined by tubal-like epithelium in sites other than the fallopian tube, is mostly described in the peritoneum and retroperitoneal lymph nodes; the main theories proposed for its pathogenesis are the shedding and implanting of tubal cells or the origin from müllerian rests. Until now, the presence of endosalpingiosis has been rarely seen outside the abdomen, mostly in the form of axillary lymph node involvement. We present the case of a paravertebral cystic lesion arising in the posterior mediastinum showing tubal-like microscopic features resembling endosalpingiosis. A 51-year-old woman presented with uterine bleeding; curettage revealed an endometrioid carcinoma. During the work-up a paravertebral lesion was found. Imaging (Fig. 1A) showed that this was an oval lesion of 3 cm at the height of the fourth thoracic vertebra without contrast uptake, compatible with a benign rather than metastatic lesion. The patient underwent hysterectomy and bilateral annexectomy, where a stage IA grade 1 endometrioid carcinoma was found. Ovaries comported few follicular cysts. No evidence of endometriosis or endosalpingiosis was found during this operation. Six months later, the patient underwent excision of the mediastinal lesion through video-assisted thoracoscopy (VATS). During the procedure, a cystic lesion supplied by large vessels and nerves was found and thus a clinical diagnosis of lymphangioma was made. The lesion was completely excised. Microscopic examination (Fig. 1B,C) revealed a cyst lined by cuboidal or cylindrical, often ciliated cells. No atypia or mitoses were found. The wall contained smooth muscle fibres. No particular stroma, siderophages or haemorrhage were found. Immunohistochemically, lining cells were positive for pan-keratins, CK7, WT1, oestrogen and progesterone receptors (Fig. 1D,E), and negative for CK20, calretinin, D2-40, and CD31. These findings were compatible with a 'müllerian' morphology and immunophenotype, mostly of the type of endosalpingiosis, in this case forming a unilocular cystic lesion resembling a hydrosalpinx. In 2005, Hattori described a paravertebral lesion lined by Müllerian-type epithelium3Hattori H. Ciliated cyst of probable mullerian origin arising in the posterior mediastinum.Virchows Arch. 2005; 446: 82-84Crossref PubMed Scopus (36) Google Scholar and shortly after he reported three such cysts in a series of 19 paravertebral cysts.4Hattori H. High prevalence of estrogen and progesterone receptor expression in mediastinal cysts situated in the posterior mediastinum.Chest. 2005; 128: 3388-3390Crossref PubMed Scopus (24) Google Scholar Two years later, Thomas-de-Montpréville and Dulmet reported another nine such cases, which comprised 5.5% of 163 non-neoplastic mediastinal cysts operated.5Thomas-de-Montpréville V. Dulmet E. Cysts of the posterior mediastinum showing müllerian differentiation (Hattori's cysts).Ann Diagn Pathol. 2007; 11: 417-420Abstract Full Text Full Text PDF PubMed Scopus (24) Google Scholar Since then, few other cases have been reported and, between them, to the best of our understanding, Batt and colleagues6Batt R.E. Mhawech-Fauceglia P. Odunsi K. Yeh J. Pathogenesis of mediastinal paravertebral müllerian cysts of Hattori: Developmental endosalpingiosis-müllerianosis.Int J Gynecol Pathol. 2010; 29: 546-551Crossref PubMed Scopus (26) Google Scholar are those speaking of endosalpingiosis in this context. Although endosalpingiosis has been found in numerous sites, gynaecological pathology reports do not cite paravertebral cysts as part of it. Endosalpingiosis, in contrast to endometriosis, is actually a disease of post-menopausal women with a mean age at diagnosis of about 52 years.7Esselen K. Terry K. Samuel A. et al.Endosalpingiosis: more than just an incidental finding at the time of gynecologic surgery?.Gynecol Oncol. 2016; 142: 255-260Abstract Full Text Full Text PDF PubMed Scopus (28) Google Scholar, 8Prentice L. Stewart A. Mohiuddin S. Johnson N.P. What is endosalpingiosis?.Fertil Steril. 2012; 98: 942-947Abstract Full Text Full Text PDF PubMed Scopus (44) Google Scholar Its prevalence in a gynaecological surgery cohort was 1.4%. Endosalpingiosis was found at the following locations in a series of 110 cases: ovary (36.8%), fallopian tube (21.3%), omentum (8.8%), and uterus (6.6%); other sites included peritoneum, abdominal wall, lymph nodes (pelvic, parametrial, obturator, external iliac), cervix, bladder, appendix, small bowel and sigmoid colon.8Prentice L. Stewart A. Mohiuddin S. Johnson N.P. What is endosalpingiosis?.Fertil Steril. 2012; 98: 942-947Abstract Full Text Full Text PDF PubMed Scopus (44) Google Scholar Two main theories are proposed to explain its pathogenesis: an origin from the secondary müllerian system, meaning ectopic tissue derived during embryological development (developmental or choristomatouts theory) or an acquired origin either from detached and implanted tubal epithelium or by coelomic metaplasia (acquired or implantational theory).9Russell P. Anderson L. Evidence for lymphatic pathogenesis of endosalpingiosis: the more things change, the more they stay the same.Pathology. 2016; 48: 1-4Abstract Full Text Full Text PDF PubMed Scopus (6) Google Scholar Lymphatic spread should also be considered, especially in cases where there is 'traumatic' condition, like florid chronic salpingitis.9Russell P. Anderson L. Evidence for lymphatic pathogenesis of endosalpingiosis: the more things change, the more they stay the same.Pathology. 2016; 48: 1-4Abstract Full Text Full Text PDF PubMed Scopus (6) Google Scholar, 10Russell P. van der Griend R. Anderson L. Yu B. O'Toole S. Simcock B. Evidence for lymphatic pathogenesis of endosalpingiosis.Pathology. 2016; 48: 72-103Abstract Full Text Full Text PDF PubMed Scopus (7) Google Scholar A traumatic origin of ectopic tissue in the pleura is seen mostly with ruptured spleen,11Biron-Schneider A. Clemenson A. Tiffet O. Perrot J. Peoc'h M. Gentil-Perret A. Thoracic splenosis mimicking pleural and pulmonary metastasis.Ann Pathol. 2010; 30: 382-385Crossref Scopus (5) Google Scholar but in mediastinal müllerian cysts, clinicopathological findings are repeated in every case, and this could not be explained by a post-traumatic origin. Similarly to the present case, thoracic cysts showing müllerian differentiation have been reported only in women, almost always in their fifties, with the exception of an 18-year-old woman.3Hattori H. Ciliated cyst of probable mullerian origin arising in the posterior mediastinum.Virchows Arch. 2005; 446: 82-84Crossref PubMed Scopus (36) Google Scholar They are usually accidental findings, with rare cases presenting with cough or pain. They are always of paravertebral localisation, arising between the third end eighth thoracic vertebra, right or left, and they measure from 1.5 to 5 cm. Preoperatively, they are mostly diagnosed as benign neural tumours or bronchogenic cysts. Microscopically, they show tubal differentiation with ciliated cells lying over a thin muscular wall. The presence of a muscular wall and of feeding vessels, findings quite uncharacteristic of 'implantation' endosalpingiosis, as well as the constant localisation of these cysts, are highly suggestive of a developmental origin. They could represent a remnant of the anlage of the fallopian tube which is found at the level of the third to fifth thoracic vertebra during embryonic development.6Batt R.E. Mhawech-Fauceglia P. Odunsi K. Yeh J. Pathogenesis of mediastinal paravertebral müllerian cysts of Hattori: Developmental endosalpingiosis-müllerianosis.Int J Gynecol Pathol. 2010; 29: 546-551Crossref PubMed Scopus (26) Google Scholar, 12Ludwig K.S. The Mayer-Rokitansky-Kuster syndrome. An analysis of its morphology and embryology. Part I: Morphology.Arch Gynecol Obstet. 1998; 262: 1-26Crossref PubMed Scopus (44) Google Scholar In conclusion, paravertebral cysts of müllerian differentiation are rarely reported, probably because they are under-diagnosed. They are always of tubal differentiation, of paravertebral thoracic localisation and a finding of menopause. Their consistent localisation and clininal, macroscopic and microscopic appearance suggest a developmental/embryological origin, rather than an acquired form of endosalpingiosis. The authors state that there are no conflicts of interest to disclose.