Comparative immunohistochemical study of deep infiltrating endometriosis, lymph node endometriosis and atypical ovarian endometriosis including description of a perineural invasion

Endometriosis is a common, oestrogen-dependent, in- flammatory disease associated with infertility and pelvic pain. It is defined as the presence of endometrium outside the uterus1. The endometriosis classification of the revised Amer- ican Society for reproductive Medicine (rASRM) differ - entiates between deep infiltratitng endometriosis (DIE), peritoneal endometriosis and ovarian endometriosis 2. Diagnostic criteria for DIE have been proposed - either they are lesions that affect the intestine, urinary bladder, ureter, vagina or uterosacral ligaments or lesions that ex- tend more than 5 mm below the peritoneal surface3. DIE represents a subset of endometriosis with an ill-defined set of symptoms and lack of specific diagnostic criteria which frequently cause significant delays in treatment4. Endometriosis is traditionally considered a benign dis- ease. However, some features that are typical of malignant disease have been identified, including metastatic spread to lymph nodes and distant organs, abnormal morphol - ogy, migration along the nerve bundles, deregulated cell growth, cellular invasion, neoangiogenesis, DNA aneu - ploidy and loss of heterozygosity5. Lymph node involvement due to endometriosis is uncommon, but it has been documented as a frequent event (25-42%) in bowel, rectosigmoid and rectovaginal endometriosis 6-8. Microscopically, two stages of lymph node endometriosis (LNE) are distinguishable – isolated endometriotic-like cells (early stage) and metastatic endo- metriotic lesions (late stage) (ref.7). A variety of hyperplastic and atypically hyperplastic changes similar to those occurring in the eutopic endo - metrium may appear in endometriotic lesions 9. In view of this finding, the term atypical endometriosis was in - troduced. The pathogenesis of endometriosis is not entirely clear. However, oestrogen and progesterone imbalance seems to play a key role in the migration, proliferation and infiltration of the peritoneum by endometrial cells10. Hormonal imbalance in endometriosis also represents a target in the treatment of the disease11. The main goal of this study was to determine and compare the status of both oestrogen and progesterone receptors in a tissue of DIE, LNE and atypical ovarian endometriosis using immunohistochemical methods. The possible relationship between endometriosis and protein p53 was also investigated. The p53 molecule is a tumor suppressor that pre - vents the outgrowth of aberrant cells by inducing cell cycle arrest, DNA repair, or programmed cell death12. Its commonly mutated in several human cancers, but p53 accumulation may also occur in non-tumour cells under various forms of stress (e.g. in a variety of autoimmune diseases) (ref.13). Staining is associated with aneuploidy, increased S phase fraction and genetic instability14.

Study group and tissue specimens A total of 40 patients with DIE were included in our study, the mean age was 33.1 years (range 22.0 - 47.0); ten patients were cigarette smokers (25%) (Table 1). All patients underwent surgery for DIE at the centre for the treatment of endometriosis at Znojmo Hospital between 2016-2018. The nature of the surgery varied widely de - pending on the extent of the disease. A minimally inva - sive laparoscopic approach was always preferred, but in 10 cases, conversion to laparotomy was neccesary. For classification of endometriosis rASRM score and ENZIAN classification were used. All patients were premenopausal, knowledge of the date of the last menstrual period and the length of the menstrual cycle was derived from the available anam - nestic data. Nine of the 40 patients received hormonal therapy at the time of surgery, 8 patients took progestin- containing medication (treatment for endometriosis, dysfunctional uterine bleeding or birth control) and 1 patient took gonadotropin releasing hormone analogue (GnRH analogue; treatment for endometriosis). The rest had not received hormonal therapy for at lest 4 months before surgery. Histological examination confirmed a diagnosis of DIE in all selected cases. Based on histopathological analysis of resected specimens, the cases were divided into 2 groups: group 1 - lymph node endometriosis (cases with lymph node involvement; n=12) and group 2 – deep infiltrating endometriosis (cases without lymph node involvement; n=28). Cases with florid fibrosis, inflam - matory or regressive changes were not included in the study. Five cases of atypical ovarian endometriosis were also included in our study. As a control group, eutopic endometrium of adenomyosis- and endometriosis-free women was used (n=16). Uterine samples were obtained from normally cycling women undergoing hysterectomy for cervical intraepithelial neoplasia (n=8), leiomyomata (n=4) or chronic pelvic pain (n=4). The diagnosis of endometriosis was based on the presence of both endometriotic glands and stroma while diagnosis of atypical endometriosis was based on histopathological criteria reported by LaGrenade and Silverberg15 and Czernobilsky and Morris 16. These crite- ria include eosinophilic cytoplasm, large hyperchromatic or pale nuclei with moderate to marked pleomorphism, an increased nuclear to cytoplasmic ratio, cellular crowd- ing and stratification or tufting. Three or more of these criteria were present in all cases (Fig. 1a). Hematoxylin and eosin stained sections of endome - trial samples were dated by the criteria described by Noyes et al.17. As in eutopic endometrium, the histologic appearance of the endometrial epithelium and stroma in endometriotic lesions was classified as early, mid and late proliferative and early, mid and late secretory. Some cases with early proliferative morphology have partly acquired an inactive appearance (characterised by small glands lined by low columnar epithelium with minimal mitotic activity). Biomed Pap Med Fac Univ Palacky Olomouc Czech Repub. 2021 Mar; 165(1):69-79. 71 Table 1. Descriptive statistics of study population including pathologic features of deep infiltrating endometriosis and immuno- histochemical analysis of oestrogen and progesterone receptor expression in deep infiltrating endometriosis (group 2). Case Age Medications Site of endo- metriosis Lymph node involvement Day of cycle Histologic appearance of deep infiltrating endometriosis ER expression* PR expression* 1 30 GnRH analogue Bladder + 58 EP 100 100 2† 37 None Vagina - 10 MP 100 100 3 45 Progestins Sigmoideum - 5 A 90 0 4 28 None USL - 24 LS 90 15 5† 35 None Sigmoideum + 11 MP 100 100 6 40 Progestins USL - 37 I 100 100 7 29 None Rectum + 10 MP 100 100 8† 32 None Rectum - 24 MP 100 100 9 33 None USL - 43 EP 100 100 10 22 Progestins USL - 11 I 100 100 11 27 Progestins USL - 30 I 100 100 12 43 Progestins USL - 31 I 100 100 13 29 None USL - 14 I 40 70 14 41 None USL - 22 I 90 80 15† 34 None USL + 8 EP 100 100 16 26 None USL + 8 I 100 100 17 24 None Bladder + 6 EP 90 70 18 27 None USL - 14 I 100 100 19 41 None USL - 15 I 90 90 20† 29 None Vagina - 18 EP 100 100 21 25 None Vagina - 20 LP 100 100 22 28 None Vagina - 32 MS 80 1 23 25 None Bladder - 21 MS 90 5 24 28 Progestins Bladder - 45 A 80 1 25 35 None Vagina + 5 MP 100 100 26 42 None Bowel - Unknown C 90 5 27 25 None USL + 3 I 100 100 28 37 None USL - 14 I 100 100 29 35 None USL - 20 C 90 10 30 36 None USL - 10 MP 100 100 31 41 None USL - 7 I 100 100 32 24 None Rectum + 5 EP 100 100 33 26 Progestins Rectum + 11 I 100 100 34 35 None USL - 4 I 100 100 35 28 None USL - 5 EP 100 100 36 42 None Sigmoideum + 24 EP 90 100 37 40 None USL - 17 EP 100 100 38 31 None Rectum + 15 MP 100 100 39 47 Progestins Bladder - 48 I 100 100 40 33 None USL - 8 EP 100 100 * Percentage (%) of marker positive endometrial glandular cells. † Cases with nerve involvement by endometrios. A, atrophy with weak secretory changes; C, a combination of proliferative and secretory endometrium; EP, early proliferative; ER, oestrogen receptor; GnRH, gonadotropin releasing hormone; LP, late proliferative; LS, late scretory; MP, mid proliferative; MS, mid secretory; PR, proges- terone receptor; USL, uterosacral ligament. Routinely processed formalin fixed and paraffin em - bedded tissues were available for immunohistochemical analysis in all cases. The tissue specimens, slides, and histopathology reports were retrieved from the files of the department of pathology. All resection specimens were fixed in 10% neutral buffered formalin for 24 h. The 4 µm thick tissue sections stained with hematoxylin-eosin were reviewed by pathologist and representative samples with structures of deep infiltrating endometriosis, lymph node endometriosis and atypical ovarian endometriosis were selected for immunohistochemical analysis. The study was approved by the ethics comittee of Biomed Pap Med Fac Univ Palacky Olomouc Czech Repub. 2021 Mar; 165(1):69-79. 72 Znojmo Hospital. The requirement for patient consent was waived becasuse of the retrospective nature of this study. Immunohistochemistry Immunohistochemistry was performed for all cases, using antibodies against oestrogen receptors (clone ER SP1, Roche, RTU), progesterone receptors (clone PR 1E2, Roche, RTU), p53 (clone Bp53-11, Roche, RTU) and Ki-67 (clone 30-9, Roche, RTU). All immunohisto - chemical staining was performed on the automated immu- nostainer BenchMark GX (Ventana Medical System, Inc., Tucson, AZ, USA); ultraView Universal DAB Detection Kit was used. The primary antibodies were visualised using hydrogen peroxide substrate and 3, 3'- diaminobenzidine tetrahydro- chloride (DAB) chromogen. External positive controls were used for all antibodies (invasive breast carcinoma for ER and PR, tonsil for p53 and lymph node for Ki-67). Negative controls were prepared by incubating samples without a primary antibody. Evaluation of immunohisto- chemical results was performed using a uniform micro - scope and camera setting (Olympus BX43F microscope and Promican 3-3CC camera). Evaluation of immunostaining In the endometrial tissue, nuclear immunostaining for ER, PR, p53 and Ki-67 was regarded as positive. The percentages of marker positive cells were evaluated using a light microscope at ×200 magnification. At least five foci of eutopic/ectopic endometrium were analyzed and the percentage of marker positive cells of the covered area was determined. The intensity of immunoreaction with antibodies against ER, PR and p53 was also investigated; weak, moderately and strong staining was distinguished. Statistical analysis Standard descriptive statistics were used to summarise the data. Continuous variables were expressed as means and standard deviation, and statistically significant dif - ferences (at the level of α=0.05) between means were evaluated using the non-parametric Mann-Whitney rank sum test with continuity correction where appropriate. Fig. 1. Atypical ovarian endometriosis and immunohistochemical expression of ER, PR and p53 in atypical ovarian endometriosis. A: Atypical ovarian endometriosis consisting of cells with eosinophilic cytoplasm, large hyperchromatic or pale nuclei with mod- erate to marked pleomorphism, an increased nuclear to cytoplasmic ratio, cellular crowding, stratification and tufting. Note the absence of the acute inflammatory infiltrate (hematoxylin-eosin staining, original magnification ×200). B: Nuclear expression of ER in approximately 40% of endometrial glandular cells with only moderately staining (immunohisto - chemistry, original magnification ×200). C: Nuclear expression of PR in less than 1% of endometrial glandular cells (immunohistochemistry, original magnification ×200). D: Strong nuclear expression of p53 in approximately 28% of endometrial glandular cells (immunohistochemistry, original mag - nification ×200). Biomed Pap Med Fac Univ Palacky Olomouc Czech Repub. 2021 Mar; 165(1):69-79. 73

DIE was diagnosed in all 40 patients with incidentally removed lymph nodes. In our series, which was a total of 118 coincidentally resected lymph nodes; the endometri- otic lesions were noted in 15 of them (12.7%). Pathologic features of lymph node endometriosis (group 1) (Table 2) All cases of LNE consisted of both endometrial glan- dular and stromal elements located in the lymph node capsule or cortex. The glands were mostly separated, fo - cally cystic dilated, occasionally formed papillary struc - tures and in one case surrounded by a lesser amount of hypocelular dense fibrous stroma. No microcalcification or psammoma bodies were found. The endometriotic glands were lined with weak proliferative endometrial epithelium with mild cellular atypia (early proliferative appearance), focal ciliated metaplasia and low mitotic activity; secretory changes were not found. Only in the case of stromal fibrosis did the epithelium seem to be flattened (atrophic). Haemosiderin-laden macrophages and pseudoxanthoma cells were a common finding in the stroma (Fig. 2a-c). Pathologic features of deep infiltrating endometriosis (group 2) (Table 1) The predominant pattern of the endometriotic tissue was early proliferative (10 cases, 25%) or inactive (15 cases, 37.5%). The others were classified as mid prolifera- tive (7 cases, 17.5%), late proliferative (1 case, 2.5%), mid secretory (2 cases, 5%) and late secretory (1 case, 2.5%). In two cases, a combination of proliferative and secretory endometrium was found (5%). In 9 patients receiving hor- monal therapy, the inactive pattern was found in 7 cases, while the atrophic pattern with weak secretory changes was found in 2 cases (5%). When more than one endometriotic focus was exam- ined in the same patient, the appearance of the specimens did not differ significantly. The appearance of endometriotic tissue was synchro- nous with the dating of the menstrual cycle in 11 cases (27.5%) – early proliferative (5 cases), mid prolifera - tive (4 cases), mid secretory (1 case) and late secretory (1 case). The most common changes in DIE were ciliated (tub- al) metaplasia, cystic changes and focal endometrial hy - perplasia without atypia. Cystic changes were found in 19 cases (47.5%), endometrial hyperplasia without atypia in 2 cases (5%) and a combination of both in 6 cases (15%). In a total of 13 cases (32.5%), no structural changes were found. Endometrial hyperplasia was characterised by out- pouching and dilatation of glandular epithelium (simple hyperplasia), but with an increasing degree of architec - tural abnormality, glands become complex, branched, crowded or back-to-back (complex hyperplasia). The lin- ing cells were columnar, mostly stratified, with oval or elongated bland nuclei with smooth contours, fine even chromatine and variable mitotic activity (no more than seven mitotic figures per 10 high-power fields). Only in 1 case (2.5%), was mucinous metaplasia found. Eutopic endometrium was available in 6 cases of DIE. In 4 (67%), the histologic appearance of endometriotic tissue was synchronous with the corresponding eutopic endometrium. Perineural and intraneural invasion (PNI) was detect- Table 2. Pathologic features of lymph node endometriosis and immunohistochemical analysis of oestrogen and progesterone receptor and p53 expression in lymph node endometriosis (group 1) and atypical ovarian endometriosis. Lymph node endometriosis Atypical ovarian endometriosis Case Histologic appearance of lymh node endo metriosis ER expression* PR expression* p53 expression* Case ER expression* PR expression* p53 expression* 1 EP 100 95 1 1 40 1 20 2 EP 100 100 1 2 55 0 26 3† A 90 90 0 3 70 0 27 4 A 95 95 1 4 55 1 28 5 EP 90 90 0 5 60 0 27 6 EP 100 95 0 - - - - 7 EP 90 95 1 - - - - 8 EP 95 90 0 - - - - 9 A 90 90 1 - - - - 10† EP 100 90 1 - - - - 11 EP 95 95 0 - - - - 12 EP 100 90 0 - - - - * Percentage (%) of marker positive endometrial glandular cells. † Involvement of 2 lymph nodes by endometrios. A, atrophic; EP, early proliferative; ER, oestrogen receptor; PR, progesterone receptor. Biomed Pap Med Fac Univ Palacky Olomouc Czech Repub. 2021 Mar; 165(1):69-79. 74 Fig. 2. Immunohistochemical expression of ER, PR and p53 in lymph node endometriosis. A: Strong nuclear expression of ER in more than 90% of endometrial glandular cells. Periglandular fibrosis, haemosiderin-laden macrophages and pseudoxanthoma cells are noted (immunohistochemistry, original magnification ×200). B: Strong nuclear expression of PR in more than 90% of endometrial glandular cells (immunohistochemistry, original magnifica- tion ×200). C: Weak nuclear expression of p53 in less than 1% of endometrial glandular cells (immunohistochemistry, original magnification ×200). Fig. 3. Endometriotic neuropathy. A: Intraneural spread of endometriotic gland (visible in the lower right corner) with marked neural hypertrophy and hyperplasia (hematoxylin-eosin staining, original magnification ×100). B: Involvement of the parasympathetic ganglion of the autonomic nervous system by endometriosis. Numerous bundles of nerve fibers, cell body of the postsynaptic neuron (visible in the right half of the field) and deposits of haemosiderin are noted (hema- toxylin-eosin staining, original magnification ×100). Biomed Pap Med Fac Univ Palacky Olomouc Czech Repub. 2021 Mar; 165(1):69-79. 75 ed in a total of 5 cases (12.5%) – rectal endometriosis (1 case), rectosigmoid endometriosis (1 case), vaginal endometriosis (2 cases) and endometriosis of the utero - sacral ligament (1 case). The nerve involvement due to endometriosis occurred in the organ (bowel or vaginal) wall and also in the adjacent tissue. PNI was a multifo - cal finding with concomitant marked neural hypertrophy and hyperplasia (Fig. 3a). An interesting finding was the involvement of the parasympathetic (terminal) ganglia located in, or close to, the wall of the affected organs (Fig. 3b). Analysis of ER, PR, Ki-67 and p53 expression in the control group (Table 3) Expression of ER in glands of normal endometrium was maximal in the proliferation phase (more than 90%), slightly decreased in the early and mid secretory phase (about 80%) and lowest being in the late secretory phase (15-50%). Expression of PR was identical in the prolif - erative and early secretory phase, a significant decrease occurred in the mid secretory (less than 10%) and the late secretory phase (less than 1%; P<0.001). For both types of hormonal receptors, expression in endometrial stromal cells was high, regardless of the phase of the menstrual cycle (90%). The intensity of staining in both endometrial glands and stroma for ER and PR was moderate or strong. The proliferation index (Ki-67) in glands also showed a decrease during the cycle, the highest values being reached in the mid proliferative phase (50-90%, on aver - age 70%), lower in the late proliferation phase (1-40%, on average 22%), while a significant decrease with minimal expression was observed in mid and late secretory phase (about 1%; P<0.001). Weak nuclear expression of p53 in less than 3% in both type of cells was detected. Analysis of ER, PR, Ki-67 and p53 expression in the lymph node endometriosis (group 1) (Table 2) and deep infiltratring endometriosis (group 2) (Table 1) The results of immunohistochemical staining with an- tibodies against ER, PR, Ki-67 and p53 were somewhat similar to those in the control group. In both groups, strong nuclear expression of ER and PR in more than 90% of cells in both endometrial glands and stroma in the proliferative phase was observed (i.e. there were no differences between LNE and DIE in the proliferative phase) (Fig. 2a,b and Fig. 4a,c). Only in one case in group 2, decrease in the PR expression only in glandular cells (70%) was found. In group 2, in the mid and late secretory phase, ER expression only slightly decreased (to more than 80%) (Fig. 4b), whereas a significant decrease of PR in endometrial glands was detected (on average 7%; P<0.001) (Fig. 4c). In group 1, all cases with atrophic morphology showed expression of ER and PR in more than 90% of glandular and stromal cells. In both groups, Ki-67 expression in endometrial glands was quite variable. In the early proliferative phase it ranged between 1 and 50% (on average 34.2% for both groups), in the mid proliferative phase it ranged between 10 and 70% (on average 44% only for group 2) and in the late proliferative phase it was 10% (only group 2); no Table 3. Immunohistochemical analysis of oestrogen and progesterone receptor, p53 and Ki-67 expression in eutopic endometrium (control group). Case Day of cycle Stage of cycle ER expression* PR expression* p53 expression* Ki-67 expression* 1 6 EP 100 100 0 40 2 8 EP 100 100 2 60 3 10 MP 100 100 1 70 4 9 MP 100 95 3 50 5 9 MP 100 100 0 70 6 11 MP 100 100 3 90 7 13 LP 95 90 1 1 8 13 LP 90 90 1 40 9 12 LP 100 100 1 10 10 16 ES 90 90 1 30 11 18 ES 100 100 1 25 12 22 MS 80 10 0 1 13 20 MS 80 2 0 1 14 19 MS 75 1 0 1 15 23 LS 15 0 0 1 16 23 LS 50 1 1 1 * Percentage (%) of marker positive endometrial glandular cells. EP, early proliferative; ER, oestrogen receptor; ES, early secretory; LP, late proliferative; LS, late secretory; MP, mid proliferative; MS, mid secre- tory; PR, progesterone receptor. Biomed Pap Med Fac Univ Palacky Olomouc Czech Repub. 2021 Mar; 165(1):69-79. 76 Fig. 4. Immunohistochemical expression of ER, PR and p53 in deep infiltrating endometriosis in the mid proliferative and mid secretory phase of the cycle. A: Diffuse strong nuclear expression of ER in both endometrial glands and stroma in the mid proliferative phase of the cycle (im- munohistochemistry, original magnification ×200). B: Diffuse strong nuclear expression of ER in both endometrial glands and stroma in the mid secretory phase of the cycle (im - munohistochemistry, original magnification ×200). C: Diffuse strong nuclear expression of PR in both endometrial glands and stroma in the mid proliferative phase of the cycle (im- munohistochemistry, original magnification ×200). D: Loss of PR expression in endometrial glands but not in endometrial stroma in the mid secretory phase of the cycle (immuno- histochemistry, original magnification ×200). E: Absence of p53 expression in both endometrial glands and stroma in the mid proliferative phase of the cycle (immunohisto - chemistry, original magnification ×200). F: Absence of p53 expression in both endometrial glands and stroma in the mid secretory phase of the cycle (immunohistochem- istry, original magnification ×200). differences between LNE and DIE in the proliferative phase were found (in group 1, no secretory changes were found). As in the control group, a significant decrease in the secretory phase in group 2 was observed (less than 15%; P<0.001). In both groups, weak and sporadic nuclear p53 expres- sion in less than 1% in endometrial glands and stroma was detected (Fig. 2c and Fig. 4e,f). In group 2, the immunophenotype of inactive appear- ing glands was as follows: strong nuclear expression of Biomed Pap Med Fac Univ Palacky Olomouc Czech Repub. 2021 Mar; 165(1):69-79. 77 both ER and PR in more than 90% of cells, minimal Ki-67 expression (less than 1%) and absence of p53 expression (0%). In one case, significant decrease in ER expression (40%), and in two cases a decrease in PR expression (70 and 80%) was observed. In group 2, in the two cases of atrophic pattern with secretory changes, ER expression was observed in more than 80% of glandular cells, while PR expression was al - most lost (less than 1%). In endometrial stromal cells, high expression of both hormonal receptors was detected (more than 80%). In the remaining two cases in group 2 (a combination of proliferative and secretory endometrium) ,,secretory pattern‘‘ of hormonal receptor expression in endometri- otic glands was observed (high ER expression and very low PR expression). Analysis of ER, PR, p53 and Ki-67 in atypical ovarian endometriosis (Table 2) A different immunophenotype in all 5 cases of atypi- cal ovarian endometriosis was found. Compared to the control group and groups 1 and 2, differences were related to ER, PR and p53. Expression of ER was on average 56% (range 40-70%) (Fig. 1b), whereas expression of PR was less than 1% (Fig. 1c). Differences for both hormonal receptors were statistically significant (P<0.001); the stain- ing was moderately or strong. Only in the lesions of atypical ovarian endometrio - sis was a strong nuclear p53 expression detected. The number of positive glandular cells ranged between 20 and 28% (on average 26%) (Fig. 1d). Compared to the control group and groups 1 and 2, differences reached statistical significance (P<0.001). The Ki-67 values ranged between 5-50% (on average 32%). Analysis of ER, PR, p53 and Ki-67 in simultaneously sampled eutopic endometrium Eutopic endometrium was available in 6 cases of DIE. When the immunophenotypes of simultaneously sampled eutopic endometrium and endometriotic foci were com - pared, there were 3 cases where differences only in the expression of Ki-67 were found; a significant decrease in endometriotic foci was found (17% vs. 67%).

In our study, we retrospectively reviewed a series of 120 cases of ovarian endometriomas; the diagnosis of atypical endometriosis was determined in 5 (4.2%). In all these cases, the criteria described by LaGrenade and Silverberg15 and Czernobilsky and Morris 16 were strictly followed for diagnosis. However, despite these histological criteria, severe epithelial atypia may be of reactive origin and the distinction is therefore problematic and limited. Some studies concluded that atypical endometriosis is not a result of reactive inflammation but rather possesses a precancerous potential because epithelial atypia occur irrespective of severe stromal inflammation, and are con- sidered the precursor lesion for endometriosis associated carcinomas (most commonly clear cell or endometrioid). None of our cases of atypical endometriosis exhibited more than a scant inflammatory infiltrate (mostly rep - resented by rare intraepithelial neutrophils). The endo - metrioid epithelium was focally denuded and, in some cases, it was problematic to distinguish between moderate (reactive) and severe celular atypia. In these cases, we used a simple immunohistochemical panel with antibod- ies against ER, PR, p53 and Ki-67. The immunophenotype of endometriotic lesions with severe celular atypia was as follows: significantly higher strong p53 expression and lower ER and PR expression with only moderate staining. The immunophenotype of cases with reactive (moderate) atypia as well as the Ki-67 values were comparable to non- atypical endometriotic lesions. Based on these findings, all ovarian endometriotic lesions with such an immuno - phenotype and severe celular atypia were classified as atypical endometriosis. A significant decrease in receptor content indicates a greater degree of autonomy of atypi - cal endometriotic lesions. These findings, together with a strong and significantly higher p53 expression support the preneoplastic potential of the atypical endometriotic tissue. Although the diagnosis of atypical endometriosis should be based primarily on morphology, finding of such an immunophenotype (in the appropriate morphological context) is a very useful ancillary method. Diagnosis of atypical endometriosis should be fol - lowed by careful long-term observation of the patient; the reason is the possible subsequent development of neoplasia. When searching the literature, the incidence of atypical endometriosis in ovarian cancer ranges from 4.4 to 22.8% and without neoplasm involvement from 1.7 to 32.3% (ref.18-21). At our hospital, the frequency of atypi- cal endometriosis was 4.2% between 2016-2018; so far no neoplasm has developed in any of these cases. The possibility of malignant transformation of endo - metriosis was originally mentioned by Samson in 1925 (ref.22) and later by Scott in 1953 (ref. 23). However, the association between atypical endometriosis and ovarian neoplasm was originally described by LaGrenade and Silverberg in 1988 - clear cell carcinoma and endometri- oid carcinoma have been found in continuity with atypical ovarian endometriosis15. A number of studies immunohistochemically analyzed the ER and PR expression in endometrial tissue. In the study of Bur et al., it was conluded that ER expression in most endometriotic lesions corresponds to their mor - phological appearance 24. These results are fully consis - tent with the results of the presented study. In another study, Lessey et al. found that receptor content in endo - metriotic tissue was more heterogeneous, unlike eutopic endometrium, and did not undergo predictable changes in response to endogenous hormones. The authors also described a significant decrease in ER and PR in both the glands and stroma in the endometriotic tissue of pa - tients treated with hormonal therapy relative to untreated patients25. In our series, 8 women took progestins. In 2 of them progestin-mediated morphologic changes in en - dometriotic tissue were observed (atrophic pattern with Biomed Pap Med Fac Univ Palacky Olomouc Czech Repub. 2021 Mar; 165(1):69-79. 78 weak secretory changes). In these two cases, loss of PR expression (only in glands) and a slight decrease in ER ex- pression (in both stroma and glands) were noted. The his- tologic appearance of the remaining 6 cases was inactive. In general, the morphological appearance of the eutopic endometrium and endometriotic tissue after progestin treatment is influenced by several factors, in particular the dose and duration of hormonal treatment, as well as the status of the endometrium26. Some studies describe an overall reduction in PR and ER expression in endometriotic lesions relative to eutopic endometrium27,28. We observed this finding only in isolated cases. PR expression decreases significantly in the mid and late secretory phase in both endometriotic tissue and eutopic endometrium. Lower concentrations of PR receptors were observed in only 2 cases of DIE with early proliferative and inactive appearance, but not in LNE. In terms of oestrogen receptors, significant dif - ferences were found in only one case of DIE with inactive appearance, compared to LNE, where all cases showed strong ER expression in more than 90% of both glandular and stromal cells. In our study, considerable attention was paid to the analysis of PNI. We found a total of 5 cases of the nerve involvement due to endometriosis. Regardless of the anatomical location of PNI, significant changes in the af- fected nerve tissue were detected. We introduced the term ,,endometriotic neuropathy“, which is defined by distinct neuronal hypertrophy and hyperplasia with concomitant involvement of the ganglia of the autonomic nervous sys- tem (ANS) located in the wall of the affected organs as well as in the adjacent fibroadipose tissue. These ganglia, consisting of a population of both Schwann and ganglion cells, were easily detectable on the tissue sections stained with hematoxylin-eosin. Immunohistochemically, neural- appearing bundles and ganglia were strongly positive for S100 protein and neuron-specific enolase (NSE). Regarding the organization of the ANS in the ab - dominopelvic organs, sympathetic presynaptic fibers lead from the lower thoracic and upper lumbar segments of the spinal cord via abdominopelvic splanchnic nerves to the vertebral and prevertebral ganglia, where postsynaptic neurons have their cell bodies. Parasymphatetic presyn - aptic neurons send axons from the sacral segments of the spinal cord to visceral ganglia that contain bodies of the postsynaptic neurons and are generally located in the walls of the organs, such as the ganglia of the submuco - sal (Meissner’s) plexus, the deep submucosal (Henle’s) plexus and the myenteric (Auerbach’s) plexus in the ali- mentary canal29. It follows from the previous comments that, for all five of our cases, the ganglia affected by en - dometriosis are part of the parasympathetic branch of the ANS. In general, the autonomic nervous system, which is subdivided into a sympathetic, parasymphatetic and enter- ic divisions, conducts impulses to smooth muscle, cardiac muscle, and glandular epithelium. The enteric division of the ANS controls motility, exocrine and endocrine secre- tions, and blood flow through the gastrointestinal tract. It also regulates immunologic and inflammatory processes. In addition to muscle fibers and the neural network, a population of the interstitial cells of Cajal (ICC) are criti- cal for bowel motility and their function is modulated by both sympathetic and parasympathetic neural inputs 30. When nerve fibers and ganglia of the ANS are affected by endometriosis, all of these functions can be impaired. As mentioned above, in our study PNI was a multifocal finding detected in the areas of macroscopically apparent lesions but also outside the confines of these areas. PNI allows the spread of endometriosis over a large distance and thus, PNI can make R0 resection of endometriotic lesions more difficult. These facts could be one of the causes of impaired functions of the affected organs after debulking surgery with macroscopic negative resection margins as well as pain symptomatology in macroscopic inapparent endometriotic lesions. Coversely, the absence of PNI could be the cause of the well-known situation of extensive organ involvement due to endometriosis and no pain at all.

From a histological and immunohistochemical point of view, deep infiltrating endometriosis and lymph node endometriosis appear to represent the same entity. For the first time, a simple immunohistochemical panel with antibodies against ER, PR and p53 useful in diagnosing atypical endometriosis has been described. The marked endometriosis-associated neural changes (endometriotic neuropathy) could be one of the causes of impaired functions of the affected organs after debulking surgery with macroscopic negative resection margins as well as pain symptomatology in macroscopic inapparent endometriotic lesions. ABBREVATIONS ANS, Autonomic nervous system; A, Atrophy; C, a combination of proliferative and secretory endometrium; DAB, Diaminobenzidine tetrahydrochloride; DIE, Deep infiltrating endometriosis; DNA, Deoxyribonucleic acid; EP, Early proliferative; ER, Oestrogen receptors; GnRH, Gonadotropin releasing hormone; ICC, Interstitial cells of Cajal; LNE, Lymph node endometriosis; LP, Late prolifer- ative; LS, Late secretory; MP, Mid proliferative; MS, Mid secretory; NSE, Neuron-specific enolase; PNI, Perineural invasion; PR, Progesterone receptors; p53, Protein 53; rASRM, Revised American Society for reproductive medi- cine; RTU, Ready to use; USL, Uterosacral ligament. Author contributions: JL: project development, manu - script writing; RCH: manuscript writing, critically revis- ing the article; LF: statistical analysis, manuscript editing; PK, DL: manuscript writing, description of the histopa - thology findings, literature (PubMed) search. Conflict of interest statement: The authors state that there Biomed Pap Med Fac Univ Palacky Olomouc Czech Repub. 2021 Mar; 165(1):69-79. 79 are no conflicts of interest regarding the publication of this article.

1. Johnson NP , Hummelshoj L. Consensus on current manage - ment of endometriosis. World Endometriosis Society Montpellier Consortium. Hum Reprod 2013;28(6):1552-68. 2. Schenken RS, Guzick DS. Revised endometriosis classification: 1996. Fertil Steril 1997;67(5):815-16. 3. Johnson NP , Hummelshoj L, Adamson GD, Keckstein J, Taylor HS, Abrao MS, Bush D, Kiesel L, Tamimi R, Sharpe-Timms KL, Rombauts L, Giudice LC. World Endometriosis Society consensus on the clas - sification of endometriosis. World Endometriosis Society Sao Paulo Consortium. Hum Reprod 2017;32(2):315-24. 4. Lindheim SR, Glenn T, Gagneux P , Maxwell RA, Yaklic JL, Findley AD, Bhagavath B. Current Challenges in the Diagnosis of Deep Infiltrating Endometriosis. Androl Gynecol 2018. Curr Res 6:1. 5. Thomas EJ, Campbell IG. Molecular genetic defects in endometriosis. Gynecol Obstet Incest 2000;50 Suppl 1:44-50. 6. Insabato L, Pettinato G. Endometriosis of the bowel with lymph node involvement. A report of three cases and review of the literature. Pathol Res Pract 1996;192(9):957-61. 7. Mechsner S, Weichbrodt M, Riedlinger WF, Kaufmann AM, Schneider A, Köhler C. Immunohistochemical evaluation of endometriotic le- sions and disseminated endometriosis-like cells in incidental lymph nodes of patients with endometriosis. Fertil Steril 2010;94(2):457-63. 8. Noël JC, Chapron C, Fayt I, Anaf V. Lymph node involvement and lymphovascular invasion in deep infiltrating rectosigmoid endo - metriosis. Fertil Steril 2008;89(5):1069-72. 9. McCluggage WG, Bryson C, Lamki H, Boyle DD. Benign, border - line, and malignant endometrioid neoplasia arising in endome - triosis in association with tamoxifen therapy. Int J Gynecol Pathol 2000;19(3):276-9. 10. Kitawaki J, Kado N, Ishihara H, Koshiba H, Kitaoka Y, Honjo H. Endometriosis: the pathophysiology as an estrogen-dependent disease. J Steroid Biochem Mol Biol 2002;83(1-5):149-55. 11. Giudice LC. Clinical practice. Endometriosis. N Engl J Med. 2010;362(25):2389-98. 12. Ozaki T, Nakagawara A. p53: the attractive tumor suppressor in the cancer research field. J Biomed Biotechnol 2011;2011:603925. 13. Yan HL, Xue G, Mei Q, Wang YZ, Ding FX, Liu MF, Lu MH, Tang Y, Yu HY, Sun SH. Repression of the miR-17-92 cluster by p53 has an important function in hypoxia-induced apoptosis. EMBO J 2009;28(18):2719- 32. 14. Engeland K. Cell cycle arrest through indirect transcriptional repres- sion by p53: I have a DREAM. Cell Death Differ 2018;25(1):114-32. 15. LaGrenade A, Silverberg SG. Ovarian tumors associated with atypical endometriosis. Hum Pathol 1988;19(9):1080-84. 16. Czernobilsky B, Morris WJ. A histologic study of ovarian endome - triosis with emphasis on hyperplastic and atypical changes. Obstet Gynecol 1979;53(3):318-23. 17. Noyes FW, Hertig AT, Rock J. Dating the endometrial biopsy. Fertil Steril 1950;1:3-25. 18. Fukunaga M, Nomura K, Ishikawa E, Ushigome S. Ovarian atypical endometriosis: its close association with malignant epithelial tu - mors. Histopathology 1997;30(3):249-55. 19. Seidman JD. Prognostic importance of hyperplasia and atypia in endometriosis. Int J Gynecol Pathol 1996;15(1):1-9. 20. Bayramoglu H, Duzcan E. Atypical epithelial change and mutant p53 gene expression in ovarian endometriosis. Pathol Oncol Res 2001;7:33-8. 21. Oral E, IIvan S, Tustas E, Korbeyli B, Bese T, Demirkiran F, Arvas M, Kosebay D. Prevalence of endometriosis in malignant epithelial ovary tumours. Eur J Obstet Gynecol Reprod Biol 2003;109(1):97-101. 22. Sampson JA. Endometrial carcinoma of the ovary arising in endo - metrial tissue in that organ. Arch Surg 1925;10:1. 23. Scott RB. Malignant changes in endometriosis. Obstet Gynecol 1953;2:283. 24. Bur ME, Greene GL, Press MF. Estrogen receptor localization in for - malin-fixed, paraffin-embedded endometrium and endometriotic tissues. Int J Gynecol Pathol 1987;6(2):140-51. 25. Lessey BA, Metzger DA, Haney AF, McCarty KS Jr. Immuno - histochemical analysis of estrogen and progesterone receptors in endometriosis: comparison with normal endometrium during the menstrual cycle and the effect of medical therapy. Fertil Steril 1989;51(3):409-15. 26. Deligdisch L. Hormonal pathology of the endometrium. Mod Pathol 2000;13(3):285-94. 27. Attia GR, Zeitoun K, Edwards D, Johns A, Carr BR, Bulun SE. Progesterone receptor isoform A but not B is expressed in endo - metriosis. J Clin Endocrinol Metab 2000;85(8):2897-902. 28. Burney RO, Talbi S, Hamilton AE, Vo KC, Nyegaard M, Nezhat CR, Lessey BA, Giudice LC. Gene expression analysis of endometrium reveals progesterone resistance and candidate susceptibility genes in women with endometriosis. Endokrinology 2007;148(8):3814-26. 29. Ross MH, Pawlina W. Histology: A text and atlas with correlated cell and molecular biology. 5th ed. Philadelphia: Lippincott Williams and Wilkins; 2005. 30. Karemaker JM. An introduction into autonomic nervous function. Physiol Meas 2017;38(5):R89-R118.