SEARCH

SEARCH BY CITATION

Keywords:

  • fallopian tubes;
  • lipofuscin-ceroid;
  • ovarian endometriosis;
  • pseudoxanthomatous salpingitis;
  • xanthogranulomatous salpingitis

ABSTRACT

  1. Top of page
  2. ABSTRACT
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. REFERENCES

Background and aim: We describe two cases of pseudoxanthomatous salpingitis (PXS), an uncommon entity, often unknown for clinicians and pathologists. We review the literature on this topic and perform immunohistochemistry in order to study its pathogenesis. Methods: Both cases are studied by histochemical and immnunohistochemical methods with a wide panel of antibodies. Results: One case showed bilateral PXS, ovarian endometriosis and an adenomatoid tumor in the right tube. The second case showed a unilateral focus of tubal endometriosis. The first case displayed pigmented histiocytes, positive with periodic acid-Schiff (PAS), PAS with diastase digestion and focally with Fontana-Masson stains. Abundant hemosiderin granules were present in stromal subepithelial cells but no in histiocytes. In the second case Fontana-Masson and Ziehl-Neelsen methods stained diffusely the histiocytes. Perls stain was positive in some stromal cells and in numerous granular macrophages. The lymphocytic infiltrates presented mainly T cells with different proportions of CD4 and CD8 lymphocytes in each case. The tubal stroma was positive for CD10 in both cases. Conclusions: PXS is a rare condition usually associated with ovarian endometriosis. The physiopathologic way of pseudoxanthomatous histiocytes to accumulate inside the plicae of the tubes is not clear and can be diverse in different cases.


INTRODUCTION

  1. Top of page
  2. ABSTRACT
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. REFERENCES

Pseudoxanthomatous salpingitis (PXS) is an uncommon entity with less than 20 cases reported in English litterature.1–11 Multiple terms as pigmentosis tubae,1,5,6 xanthogranulomatous salpingingitis,3,7–9 PXS,2,7,9,11 pseudoxanthomatous salpingiosis4 xanthogranuloma or inflammatory pseudotumor,7 and melanosis tubae10 have been used to designate this disease. PXS is characterized by numerous histiocytes containing brown lipofuscin-ceroid pigment within the lamina propria of the tube, along with other chronic inflammatory cells.1–11 In most cases it is associated with ovarian endometriosis (OE)2,4,6–11 but other entities as pelvic inflammatory disease,5–7,9 intrauterine contraceptive device8,9 and radiotherapy1 have also been related with the development of PXS. The pathogenesis of PXS is not clear, although different ways have been proposed to explain the histiocytic infiltrate in tubarian plicae.4–10 The aim of this study is to describe two new cases of PXS in patients with OE. In both cases the histiocytic infiltrates of the tubes and the aggregates of pseudoxanthomatous cells of the OE were composed of cells with different histochemical findings. The pathogenesis of our cases of PXS seems clearly related with the presence of blood material in the lamina propria of the fallopian tubes, but the mechanism of blood spreading to the tubarian mucosa could be different in each case. Our findings show that one or more physiopathologic ways may be involved in the genesis of PXS.

METHODS

  1. Top of page
  2. ABSTRACT
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. REFERENCES

Tissues were fixed in 10% neutral-buffered formalin, embedded in paraffin, sectioned and stained with hematoxilin-eosin, periodic acid-Schiff (PAS), PAS diastase, Fontana-Masson, Giemsa and Ziehl-Neelsen methods. The immunohistochemical study was performed with the Envision Kit (Dako, Glostrup, Denmark) in a Dako Autostainer Link 48 for the antibodies cytokeratin AE1–3, Calretinin, CD3, CD4, CD8, CD10, CD20, CD34, CD43, and CD68. All of the antibodies were provided by Dako and used with the optimal work dilution. Previously, an antigen retrieval protocol was applied inside a Dako PT Link automatic immunostainer using HIER and a target retrieval solution of high or low pH (Dako) depending on the antibody.

RESULTS

  1. Top of page
  2. ABSTRACT
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. REFERENCES

Clinical findings

The first case corresponded to a 45-year-old woman with one normal pregnancy and delivery. She suffered abdominal pain and menstrual changes during several months. At the gynecological examination an abdominal-pelvic mass was found. On scan it corresponded to a right cystic adnexal mass measuring 7 cm in size. In 2008 January a hysterectomy with bilateral adnexectomy was performed with asymptomatic follow up.

The second case corresponded to a 49-year-old woman also with one normal pregnancy and delivery. She showed numerous fibroids and a total hysterectomy was made in May 2007. In October 2009 an ultrasound study disclosed a left adnexal mass of 5 cm in diameter, consistent with a cystadenoma. A bilateral adnexectomy was performed.

Pathologic findings

Case 1

An intramural uterine leiomyoma, 1 cm in size, was present at the surgical specimen. Both ovaries were multicystic and measured 3.5 × 2.5 cm in size for the left ovary and 7 × 5 cm for the right one. On cut section, both ovaries presented several cysts with irregular internal surfaces and were filled with coagulated blood material. The left tube measured 5 cm in length and 1.5 cm as maximal diameter. On cross-section, a content of hematic appearance was found inside the tube. The plicae were thickened and brown-yellow in color. The right tube measured 5.5 cm in size and was normal.

On light microscopy the ovarian cysts were lined by cylindrical or mucinous epithelium and the underlying stroma showed an endometrial appearance that was made up of spindle cells. At the areas without epithelium the cysts were lined by pigmented histiocytes with fine grayish-brown cytoplasmic granules, enmeshed with hemosiderin-laden macrophages. Scanty small lymphocytes were found inside the inflammatory infiltrate. Left fallopian tube sections revealed marked expansion of the plicae by pigmented histiocytes with light brown pigment. A moderate number of lymphocytes were also present admixed with the pigmented histiocytes (Fig. 1a). Under tubal epithelium numerous stromal cells with cytoplasmic pigment of hemosiderin appearance were found (Fig. 1b). The widest plicae were lined by thin epithelium and some of them showed areas of erosion with loss of the epithelial layer (Fig. 1a). Inside the lumina, abundant partially digested blood material was present. Focally in the mesosalpinx there were small endometriotic cysts lined by cylindrical epithelium and surrounded by scanty stroma and pigmented histiocytes. The right tube showed abundant hemosiderin in subepithelial stromal cells and scanty pigmented histiocytes in lamina propria. In the muscular layer of the right tuba an adenomatoid tumor measuring 0.7 × 0.5 cm in size was found. It showed tubules lined by cuboidal or flattened cells. PAS stain and PAS after diastase digestion were intensely positive on the pigmented histiocytes of the ovarian cysts and tubes. Perls stain showed extensive extracellular and intracellular positivity in pseudoxathomatous lesions of the OE. This technique, at the uterine tubes, was positive in scanty histiocytes but numerous subepithelial stromal cells presented granular cytoplasmic stain (Fig. 1c). Masson-Fontana method showed isolated positive granules in pigmented histiocytes of the tubes and ovaries. Ziehl-Neelsen and Giemsa were negative.

image

Figure 1. Case 1. (a) Cross section of fallopian tube, showing markedly distended plicae with pigmented histiocytes, and chronic inflammatory cells. The mucosal surface displays extensive epithelial detachment with blood material inside the lumen (hematoxylin and eosin [H&E]). (b) Granular pigment with appearance of hemosiderin on subepithelial stromal cells. Histiocytes devoid this pigment (H&E). (c) With Perls method a band pattern stain at subepithelial stromal cells is demonstrated. Pigmented histiocytes are unstained (Perls). (d) Antibody against CD4 decorates the cell membrane of the histiocytes and most of lymphocytes (immunohistochemical stain).

Download figure to PowerPoint

At the immunohistochemical study the histiocytes that infiltrated tubal mucosa were positive for CD4, CD43 and CD68. Most of the lymphoid component corresponded to T cells, stained with antibodies for CD3 and CD43, and CD4/CD8 ratio was 5/1 (Fig. 1d). Scanty, scattered B lymphocytes, CD20 positive, were also present. Strong diffuse immunostaining for CD10 was found at the stroma of the ovarian cysts and at the thin layer of stroma that surrounded the cysts of the left mesosalpinx. At both tubes the capillary of the lamina propria were diffusely positive for CD10 and this stain was focal and less strong in spindle subepithelial stromal cells. On the other hand CD34 was intensely positive on the stromal cells of the tubal mucosa but the stroma of all the endometriotic lesions was negative. Adenomatoid tumor presented a typical immunophenotype with expression of AE1-AE3 cytokeratin and calretinin.

Case 2

Hysterectomy specimen contained subserosal and intramural leiomyomas. The left ovary measured 5 × 4 cm in size and showed a unilocular cyst with irregular wall and bloody content. The left fallopian tube measured 7 cm in length, with maximal diameter of 1 cm, and showed strong adhesions to the ovarian surface. On serial sections the lumen was lightly dilated but without blood material and the plicae were thickened and yellow in color. The right ovary measured 5 × 2.5 cm in size and showed a corpus luteum and small subcortical cysts. The right tube measured 5 cm in length and was normal on macroscopic study.

On light microscopy an endometriotic cyst, lined by endometrial epithelium with underlying endometrial stroma, was found. Some areas was eroded on which the epithelium was substituted by aggregates of pigmented or hemosiderin laden histiocytes. The left tube contained extensive histiocytic infiltrates at the lamina propria. 60% of these histiocytes were pigmented type and the remaining presented coarse cytoplasmic granules of hemosiderin appearance (Fig. 2a). This hemosiderin pigment was also found in isolated subepithelial stromal cells. Scanty foamy histiocytes with vacuolated cytoplasm were present. As in the first case, abundant lymphocytes were enmeshed with the histiocytes. Focally, some neutrophils were found. On the most widened plicae the epithelium was focally thinned but erosions, ulcerated areas or hemorrhagic luminal content were not present in this case. In one section the tubal mucosa was in continuum with a focus of endometrial mucosa with tall epithelial cells that showed supra and infra nuclear vacuoles. The stroma of tubal endometriotic foci was dense with round and spindle cells (Fig. 2b). The right ovary showed follicular cysts and a corpus luteum. The right tube did not show pathological findings.

image

Figure 2. Case 2. (a) The lamina propria of the tube contains numerous pigmented histiocytes (left side) and some hemosiderin laden macrophages (right side) (hematoxylin and eosin [H&E]). (b) Focus of endometriosis at the mucosa of the tube. The endometrial mucosa is in continuum with the mucosa of the tube. On the upper right area a plica is dilated by aggregates of histiocytes (H&E). (c) The pigmented histiocytes show a strong stain with periodic acid-Schiff (PAS), whereas macrophages with hemosiderin are negative with this method (PAS). (d) Most of the lymphocytic infiltrate is composed of lymphocytes positive for CD8 (immunohistochemical stain). (e) Strong staining for CD10 at the stroma of the focus of tubal endometriosis. This stain on the stromal cells of adjacent lamina propia of the tube is lighter (immunohistochemical stain).

Download figure to PowerPoint

The pigmented histiocytes of the endometrial cyst of the left ovary and left tube stained strongly with PAS, PAS after diastase digestion and Masson-Fontana stains (Fig. 2c). Ziehl-Neelsen stain was positive only in pigmented histiocytes of the tube. In both the endometrial cyst of the left ovary and the left tube the histiocytes and some stromal cells with granular golden cytoplasmic pigment stained with Perls technique. Foamy histiocytes displayed light stain with PAS method and were negative with the other hystochemical stains. Giemsa technique was negative.

The immunohistochemical profile of the lympho-histiocytic infiltrate was similar to case 1. Most of the lymphocytes were T lymphocytes but in this case the predominant component was positive for CD8 with a CD4/CD8 ratio of 2/4 (Fig. 2d). CD10 was strongly positive at the stroma of the endometriotic cyst of the left ovary and at the focus of endometriosis of the mucosa of the tube (Fig. 2e). This marker also stained abundant capillaries and, less strongly, stromal cells of the tubal mucosa (Fig. 2e). CD34 was strongly positive at the tubal stroma but negative at the endometrial stroma.

DISCUSSION

  1. Top of page
  2. ABSTRACT
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. REFERENCES

In this study two cases with histopathological features of PXS are presented. This entity was described in 1983 by Herrera et al.1 who called it Pigmentosis tubae. Thereafter around 20 cases studied in detail have been reported although with different names: xanthogranulomatous salpingitis (XS), pseudoxanthomatous salpingiosis, PXS, xanthogranuloma, inflammatory pseudotumor, and melanosis tubae.2–11 PXS has a characteristic gross appearance. The tubal mucosa exhibits numerous small bulbous projections that microscopically correspond to plicae distended by pigmented macrophages and a lesser component of other inflammatory cells. Most of the cases of PXS are associated with endometriosis2–4,6–11 but other cases may be related to pelvic inflammatory disease,5,6,7,9 intrauterine contraceptive device,8,9 unsuccessful antibiotic therapy and radiotherapy.1,4,8 Some authors think that PXS and XS are not the same entity.7,9 For Furuya et al.7 PXS appears as a xanthogranulomatous inflammation secondary to endometriosis. On the other hand, they suggested that pelvic inflammatory disease associated with xanthogranulomatous changes should be termed as a pure XS.7 Histologic and histochemical studies show that the histiocytic components of PXS present some differences with those of XS.7,9 Other authors believe that both entities belong to a wide spectrum of changes caused by multiple etiologies and that overlaps between these entities exist.7–9

We studied two new cases of PXS in patients with OE. Microscopically each case showed the typical appearance of PXS but the lympho-histiocytic infiltrate of the plicae presented distinctive histologic, histochemical and immunohistochemical characteristics. In the first case most of the cells corresponded to pigmented histiocytes, which were stained with PAS and PAS after diastase digestion techniques, although isolated cells were also positive with Fontana-Masson stain. Most of these cells were negative with the Perls method. In the second case, pigmented histiocytes were diffusely stained with PAS, PAS after diastase digestion, Fontana-Masson and Ziehl-Neelsen stains. Moreover, numerous macrophages contained cytoplasmic coarse golden-brown pigment, which was stained with Perls technique but was negative with the other histochemical methods. In both cases the pigmented histiocytes of PXS and those of the OE showed similar histochemical features. This similar stain in pigmented histiocytes of PXS and histiocytes of the pseudoxanthomatous aggregates of OE was already described in three cases reported by Clement et al.2 It is also a well known fact that the characteristic pigmented histiocytes of endometriosis contain lipofuscin-ceroid and consequently, its histochemical profile is similar to those of the histiocytes of PSX.2,11 The histochemical differences between our two cases seem related to the type and the “age” of the pigment contained inside the histiocytes. In the first case the histochemical findings are consistent with lipofuscin, whereas in the second case the diffuse staining with Fontana-Masson and Ziehl-Neelsen methods is consistent with ceroid.7,8

In addition to the histiocytic component both cases showed a moderate lymphoid infiltrate composed basically of T lymphocytes but with different subpopulations in each case. In case 1 CD4 lymphocytes were predominant, while in case 2 CD8 lymphocytes were more abundant. This difference in lymphoid supopulations may be related to different immune responses. In case 2 it is possible that the increase of CD8 lymphocytes is a part of the response before an inflammatory-infectious process in which neutrophilic polymorphonuclear leukocytes are involved too.

The pathogenesis of PXS is not clear enough, but most authors think that it is related to the spreading of blood in the mucosa of the uterine tube.2,4,6–9 Almost all the cases of PXS are associated with OE2,4,6–10 as in the cases that we describe. The histologic findings of different authors2,11 and those of our two cases demonstrate that the aggregates of pigmented histiocytes of the ovaries appeared always in areas closed to typical endometriosis, lined by endometrial epithelium and stroma positive for CD10. Therefore, in ovary it is reasonable to think that these aggregates of histiocytes represent a physiopathologic response to the bleeding that occurs periodically in endometriosis. Nevertheless the question is: How does the blood reach the tubal mucosa? It is proposed that the repetitive hemorrhages at the cysts of OE could reach the lumen of the tubes and from here go into the mucosa.2,4,7 This way is logical, but it does not explain the low number of cases of PXS in the context of an entity as common as OE. Probably other adjuvant factors would be implicated in the pathogenesis of PXS associated to OE. On this topic Seidman et al.4 consider that the entry of blood material to the lamina propria would happen only in cases with salpingitis and loss in the integrity of tubal mucosa. This mechanism of indirect hematic diffusion could be acceptable in case 1 inasmuch as eroded areas of the epithelium and luminal hemorrhage were present.

Endometriotic foci at the mucosa or in the wall of the tube can be another mechanism involved in the pathogenesis of PXS. In this situation the origin of the hemorrhage would be intramucosal or intramural. Idrees et al.8 and Zorzi et al.10 describe cases of XS or melanosis tubae associated with fallopian mucosal endometriosis8 or parietal endometriosis,10 as it also appears in our case 2. Surprisingly in the case of Idress and in our case 2, foamy histiocytes mixed with pigmented histiocytes were present, although the significance of this finding is unknown. We also believe that in the pathogenesis of some cases of PXS the two mechanisms described above may coincide. This may also be possible in case 1 in which mucosal erosion and a focus of endometriosis at the mesosalpinx were found.

Most of the reported cases of PXS showed scanty hemosiderin. This pigment was stored in stromal cells, endothelial cells and isolated pigmented histiocytes.1,4–10 In our cases hemosiderin was abundant. In case 1 it was mainly localized in subepithelial stromal cells and in case 2 in macrophages and stromal cells. Probably the different amounts and cellular distributions of hemosiderin in the reported cases correspond to diverse steps in the development of PXS.

In summary the typical histologic findings of PXS represent the final event of the clearance of the hemorrhagic material. This process is performed by histiocytic cells and the final products of the blood degradation are stored in the cytoplasms of these cells as lipofuscin or ceroid. The source of lipofuscin-ceroid pigment is probable lipid peroxidation of lysosomes and cellular constituents. Iron-promoted lipid peroxidation may alter the lysosomal membranes and contribute to the excessive accumulation of lipofuscin-ceroid pigment in the pigmented cells of PXS.4,6–8 For Zorzi et al.10 the term “melanosis (lipofuscin) tubae,” is consistent with the mechanism in the genesis of pigmentation, which is similar to that described in melanosis coli.

On immunohistochemical study of our cases the stroma of the endometriotic lesions was intensely positive for CD10 and negative for CD34. On the other side the stroma of the tube displayed strong staining for CD34 and focal positivity for CD10. The expression of CD34 on tubal stroma is a well known phenomenon.12 Nevertheless in this location CD10 has been described as negative or positive only in small vessels.13,14 As our findings are different to those referred in the literature we studied the expression of this antigen in normal fallopian tubes (non-published results) and we found variable staining for CD10 on the stromal cells of the mucosa of many tubes, although the stain was lighter than at endometrial stroma. We believe that this result should be known by the pathologists and kept in mind for the differential diagnosis of entities as endometriosis or endosalpingiosis.

On the review of the literature we have not found any case of adenomatoid tumor and PXS in the same tube. This association found in one of our cases seems to be an incidental finding without pathogenetic relationship between both entities.

REFERENCES

  1. Top of page
  2. ABSTRACT
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. REFERENCES
  • 1
    Herrera GA, Reimann BE, Greenberg HL, Miles PA. Pigmentosis tubae, a new entity: light and electron microscopic study. Obstet Gynecol 1983; 61(3 Suppl): 80S3S.
  • 2
    Clement PB, Young RH, Scully RE. Necrotic pseudoxanthomatous nodules of ovary and peritoneum in endometriosis. Am J Surg Pathol 1988; 12: 3907.
  • 3
    Franco V, Florena AM, Guarneri G, Gargano G. Xanthogranulomatous salpingitis. Case report and review of the literature. Acta Eur Fertil 1990; 21: 1979.
  • 4
    Seidman JD, Oberer S, Bitterman P, Aisner SC. Pathogenesis of pseudoxanthomatous salpingiosis. Mod Pathol 1993; 6: 535.
  • 5
    Bolaji, II, Meehan FP. Idiopathic pigmentosis tubae. Int J Fertil Menopausal Stud 1994; 39: 869.
  • 6
    Munichor M, Kerner H, Cohen H, Iancu TC. The lipofuscin-iron association in pigmentosis tubae. Ultrastruct Pathol 1997; 21: 27380.
  • 7
    Furuya M, Murakami T, Sato O et al . Pseudoxanthomatous and xanthogranulomatous salpingitis of the fallopian tube: a report of four cases and a literature review. Int J Gynecol Pathol 2002; 21: 569.
  • 8
    Idrees M, Zakashansky K, Kalir T. Xanthogranulomatous salpingitis associated with fallopian tube mucosal endometriosis: a clue to the pathogenesis. Ann Diagn Pathol 2007; 11: 11721.
  • 9
    Kostopoulou E, Daponte A, Kallitsaris A et al . Xanthogranulomatous salpingitis: report of three cases and comparison with a case of pseudoxanthomatous salpingitis. Clin Exp Obstet Gynecol 2008; 35: 2914.
  • 10
    Zorzi MG, Pusiol T, Piscioli F. Melanosis tubae: histogenesis and appropriate terminology. Int J Gynecol Pathol 2010; 29: 24851.
  • 11
    Clement PB. The pathology of endometriosis: a survey of the many faces of a common disease emphasizing diagnostic pitfalls and unusual and newly appreciated aspects. Adv Anat Pathol 2007; 14: 24160.
  • 12
    Yamazaki K, Eyden BP. Ultrastructural and immunohistochemical studies of stromal cells in lamina propria of human fallopian tube ampullar mucosa: the recognition of ‘CD34 positive reticular network’ and its putative function for immune surveillance. J Submicrosc Cytol Pathol 1996; 28: 32537.
  • 13
    Ordi J, Romagosa C, Tavassoli FA et al . CD10 expression in epithelial tissues and tumors of the gynecologic tract: a useful marker in the diagnosis of mesonephric, trophoblastic, and clear cell tumors. Am J Surg Pathol 2003; 27: 17886.
  • 14
    Oliva E. CD10 expression in the female genital tract: does it have useful diagnostic applications? Adv Anat Pathol 2004; 11: 3105.