SKENE’S GLAND DERIVATIVES IN THE FEMALE GENITAL TRACT AND CERVICAL ADENOID BASAL CARCINOMA ARE CONSISTENTLY POSITIVE WITH PROSTATIC MARKER NKX3.1.

Abstract

Cervical ectopic prostatic tissue and vaginal tubulosquamous polyp are rare lesions which exhibit variable, and often focal, immunohistochemical expression with traditional prostatic markers (prostate specific antigen (PSA) and prostatic acid phosphatase (PSAP)). These lesions are thought to arise from periurethral Skene’s glands, the female equivalent of prostatic glands in the male. Adenoid basal carcinoma is a rare and indolent cervical neoplasm. Expression of the prostatic marker NKX3.1 in ectopic prostatic tissue and tubulosquamous polyp has been reported but no studies have examined immunoreactivity with this marker in adenoid basal carcinoma. We stained 19 cases (adenoid basal carcinoma (n=6), cervical ectopic prostatic tissue (n=11) and vaginal tubulosquamous polyp (n=3); 1 case contained both adenoid basal carcinoma and ectopic prostatic tissue) with NKX3.1. In all cases, the glandular component of these lesions exhibited diffuse nuclear immunoreactivity while normal endocervical glands were negative. PSA was positive in 4 of 9 and 0 of 3 cases of ectopic prostatic tissue and tubulosquamous polyp respectively while PSAP was positive in 3 of 4 and 2 of 2 cases of ectopic prostatic tissue and tubulosquamous polyp respectively; 3 of 5 cases of adenoid basal carcinoma tested were focally positive with PSAP and all 5 were negative with PSA. While the specificity of NKX3.1 should be investigated in future studies, positivity with this marker may be useful in diagnosing these uncommon lesions. NKX3.1 appears a more sensitive marker of ectopic prostatic tissue and tubulosquamous polyp than traditional prostatic markers and positive staining provides further support that these lesions exhibit “prostatic” differentiation and are of Skene’s gland origin. NKX3.1 and PSAP positivity in adenoid basal carcinoma raises the possibility of an association with benign glandular lesions exhibiting prostatic differentiation and we critically discuss the possible association.

INTRODUCTION

So-called ectopic prostatic tissue is a rare benign cervical lesion which morphologically is characterised by squamous and glandular elements, the latter exhibiting variable positivity with traditional prostatic markers prostate specific antigen (PSA) and prostatic acid phosphatase (PSAP) (1–3). A morphologically similar lesion exists in the vagina, tubulosquamous polyp (4–6). This polypoid lesion is also composed of squamous and glandular elements, the latter sometimes exhibiting positivity with prostatic markers. It has been speculated that cervical ectopic prostatic tissue and vaginal tubulosquamous polyp are part of a spectrum of lesions in the lower female genital tract derived from eutopic or misplaced periurethral Skene’s glands, the female equivalent of prostatic glands in the male (4,7).

In this study, we show that cervical ectopic prostatic tissue and vaginal tubulosquamous polyp exhibit consistent diffuse nuclear immunoreactivity with a relatively “new” prostatic marker NKX3.1. We also show that cervical adenoid basal carcinoma is commonly positive with this marker and speculate on a possible association between cervical ectopic prostatic tissue and adenoid basal carcinoma.

MATERIALS AND METHODS

Cases Included in Study

The cases (table 1) were derived from the institutions to which the authors are affiliated and from the referral practice of two of the authors (KJP, WGM). 19 cases were included comprising cervical adenoid basal carcinoma (n=6), cervical ectopic prostatic tissue (n=11) and vaginal tubulosquamous polyp (n=3); one case contained both adenoid basal carcinoma and ectopic prostatic tissue. As well as evaluating NKX3.1 immunoreactivity, we also recorded the results of staining with PSA and PSAP where these markers were performed. With all the markers, cases were score as negative, focal (<50% of glandular cells staining) or diffuse (≥50% of glandular cells staining).

Table 1.

Clinicopathological Features of Cases Included in Study

CASE	AGE (years)	OPERATIVE PROCEDURE	DIAGNOSIS	OTHER FINDINGS
1	33	LLETZ	Adenoid basal carcinoma and ectopic prostatic tube	HSIL
2	90	LLETZ	Adenoid basal carcinoma	Squamous cell carcinoma
3	65	LLETZ	Adenoid basal carcinoma	HSIL
4	54	LLETZ	Adenoid basal carcinoma	Squamous cell carcinoma
5	85	Cervical biopsy	Adenoid basal carcinoma	Squamous cell carcinoma
6	71	Radical hysterectomy for cervical carcinoma	Adenoid basal carcinoma	Adenosquamous carcinoma (adenoid cystic-like features)
7	45	Hysterectomy and bilateral salpingo-oophorectomy for ovarian tumour	Ectopic prostatic tissue	Tubo-ovarian high-grade serous carcinoma
8	32	LLETZ	Ectopic prostatic tissue	Microglandular hyperplasia
9	75	LLETZ	Ectopic prostatic tissue	Nil
10	29	LLETZ	Ectopic prostatic tissue	HSIL
11	65	Hysterectomy for prolapse	Ectopic prostatic tissue	Uterine adenomyosis
12	39	LLETZ	Ectopic prostatic tissue	HSIL
13	30	LLETZ	Ectopic prostatic tissue	HSIL
14	32	LLETZ	Ectopic prostatic tissue	HSIL
15	51	LLETZ	Ectopic prostatic tissue	HSIL
16	45	Hysterectomy and bilateral salpingo-oophorectomy for ovarian tumour	Ectopic prostatic tissue	Tubo-ovarian high-grade serous carcinoma
17	83	Excision of vaginal polyp/ cyst	Tubulosquamous polyp	None
18	87	Excision of vaginal polyp/ cyst	Tubulosquamous polyp	None
19	61	Excision of vaginal polyp/ cyst	Tubulosquamous polyp	None

LLETZ= large loop excision of transformation zone; HSIL= high-grade squamous intraepithelial lesion.

Immunohistochemical Methods

The NKX3.1 immunohistochemistry was performed on the Ventana Benchmark Ultra platform in Royal Derby Hospital, United Kingdom and Memorial Sloan Cancer Center, USA. In Royal Derby Hospital, the manufacturers recommended protocol was used. The antibody (antibody clone EP356) supplied by Cell Marque was ready to use and was optimally diluted by the manufacturer for use on the Ventana BenchMark Ultra in combination with Ventana detection kits and accessories. No further dilution or titration was carried out. The recommended staining protocol with ultraView detection kit included deparaffinization, antigen retrieval with Cell Conditioning 1, a primary antibody incubation time of 32 minutes at 36°C, counterstaining with Haematoxylin II for 8 minutes and post counterstaining bluing for 4 minutes. In Memorial Sloan Cancer Center, a similar protocol was used with some modifications. Normal prostatic tissue and a prostatic adenocarcinoma were used as a positive control and positive staining was identified as strong, crisp, nuclear staining with a clean background.

RESULTS

Table 1 lists the clinicopathological features of the cases included in the study.

In the cases of ectopic prostatic tissue, the patient age ranged from 29–75 years, in the cases of tubulosquamous polyp from 61–87 years and in the cases of adenoid basal carcinoma from 33 to 90 years. In all cases, the morphological features were entirely in keeping with those described in the different lesions (see DISCUSSION).

The cases of ectopic prostatic tissue were generally incidental findings in hysterectomy specimens performed for benign or malignant reasons or loop excisions performed for squamous intraepithelial lesions; in one of the hysterectomy specimens (case 11), the ectopic prostate tissue was seen macroscopically as a 13 mm nodule on the ectocervix. The cases of tubulosquamous polyp all presented as vaginal polyps or “cysts”. The adenoid basal carcinomas were either incidental findings in hysterectomy specimens performed for benign reasons or for other neoplasms or in loop excisions performed for squamous intraepithelial lesions.

Two of the cases of adenoid basal carcinoma were intimately admixed with and colonised by high-grade squamous intraepithelial lesion; the other four cases also exhibited another invasive carcinoma (3 squamous cell carcinoma, 1 adenosquamous carcinoma). In one of the cases of adenoid basal carcinoma, there was also ectopic prostatic tissue; in this case, the two lesions were separate without intermingling, the adenoid basal carcinoma being located at the transformation zone and the ectopic prostatic tissue on the ectocervix.

The immunohistochemical results are presented in table 2. In all cases of ectopic prostatic tissue and tubulosquamous polyp, there was diffuse strong nuclear immunoreactivity with NKX3.1 of the glandular elements which typically represented a minor component of the lesion largely confined to the periphery of the squamous elements. All of the adenoid basal carcinomas also exhibited diffuse strong nuclear positivity of the glandular elements. The other tumour types associated with the adenoid basal carcinoma were negative with NKX3.1 Normal cervical and vaginal squamous epithelium was generally negative with NKX3.1 but occasionally there was focal weak nuclear immunoreactivity of the basal layers of the cervical squamous epithelium or the squamous epithelium within the ectopic prostatic tissue and tubulosquamous polyps. All normal endocervical glands were totally negative.

Table 2.

Immunohistochemical Results

CASE	DIAGNOSIS	NKX3.1	PSA	PSAP
1	Adenoid basal carcinoma and ectopic prostatic tube	Diffuse	NP	NP
2	Adenoid basal carcinoma	Diffuse	Negative	Focal
3	Adenoid basal carcinoma	Diffuse	Negative	Focal
4	Adenoid basal carcinoma	Diffuse	Negative	Focal
5	Adenoid basal carcinoma	Diffuse	Negative	Negative
6	Adenoid basal carcinoma	Diffuse	Negative	Negative
7	Ectopic prostatic tissue	Diffuse	Focal	Focal
8	Ectopic prostatic tissue	Diffuse	Negative	Negative
9	Ectopic prostatic tissue	Diffuse	Focal	Focal
10	Ectopic prostatic tissue	Diffuse	Negative	NP
11	Ectopic prostatic tissue	Diffuse	Negative	NP
12	Ectopic prostatic tissue	Diffuse	NP	NP
13	Ectopic prostatic tissue	Diffuse	Focal	NP
14	Ectopic prostatic tissue	Diffuse	Negative	NP
15	Ectopic prostatic tissue	Diffuse	Negative	NP
16	Ectopic prostatic tissue	Diffuse	Focal	Focal
17	Tubulosquamous polyp	Diffuse	Negative	Diffuse
18	Tubulosquamous polyp	Diffuse	Negative	Focal
19	Tubulosquamous polyp	Diffuse	Negative	NP

NP= not performed; PSA=prostatic specific antigen; PSAP= prostatic acid phosphatase

PSA was positive in 4 of 9 and 0 of 3 cases of ectopic prostatic tissue and tubulosquamous polyp respectively while PSAP was positive in 3 of 4 and 2 of 2 cases of ectopic prostatic tissue and tubulosquamous polyp respectively (table 2); staining with these two markers was confined to the glandular elements and was usually focal with only occasional glands being positive. Three of 5 cases of adenoid basal carcinoma stained were focally positive with PSAP and negative with PSA.

Figures 1–4 illustrate representative images of the morphology and immunoreactivity with the various markers of cases included in the study.

Figure 1.

Cervical ectopic prostatic tissue composed predominantly of squamous elements but with small tubules around the periphery (A); higher power showing admixture of squamous elements and tubules (B). There is diffuse positive staining of the glandular/ tubular elements with NKX3.1 (C) and focal positive staining with PSAP (D).

Figure 4.

Cervical adenoid basal carcinoma admixed with squamous cell carcinoma (A and B). There is diffuse positive nuclear staining of the glandular component of the adenoid basal carcinoma with NKX3.1 (C) and focal positive staining with PSAP (D).

DISCUSSION

In this study, we have shown consistent positive nuclear staining with the prostatic marker NKX3.1 in cervical ectopic prostatic tissue, vaginal tubulosquamous polyp and cervical adenoid basal carcinoma. This is the first study to demonstrate positivity with this marker in adenoid basal carcinoma.

So-called ectopic prostatic tissue is uncommon within the cervix and is usually an incidental microscopic finding, although in very rare cases, a grossly visible lesion is apparent (1–3); in one of the cases in our study, the lesion was identified grossly as a small nodule on the ectocervix. The “prostatic” tissue may be present superficially or deep within the cervical stroma and is usually located predominantly on the ectocervix rather than at the transformation zone, although sometimes the latter is involved. While there is debate as to whether this represents a developmental anomaly or a metaplasia of endocervical glands, the typical ectocervical location suggests a developmental anomaly. The typical morphological appearance is of rounded aggregates of epithelial cells comprised of both glandular/ tubular and squamous elements, the latter often predominating. Tubules lined by bland epithelium are typically present around the periphery of the squamous elements and often the tubular elements, at least focally, have a double cell layer of luminal and basal cells. Cribriform and papillary patterns are common. Although the glandular/ tubular element may be positive with PSA and PSAP, this is not always the case, as shown in our study. Some examples are negative with one or both markers and often positive staining is focal rather than diffuse. PSAP is more likely to be positive than PSA (1–3); p63 and other basal markers highlight the presence of a basal cell layer within the glandular or tubular elements.

Vaginal tubulosquamous polyp, first described in a study co-authored by one of us (WGM) in 2007 (4), is a polypoid lesion which is morphologically similar to cervical ectopic prostatic tissue and may exhibit positive immunohistochemical staining with prostatic markers; in the original description of 10 cases, the tubules were positive with PSAP in 3 of 4 and with PSA in 2 of 4 cases tested (4). Uncommon findings include the presence of sebaceous glands, basaloid formations resembling hair follicle structures and a microglandular proliferation (3,4–6). Occasionally, similar microscopic lesions are seen in the vulva (3,5). It has been speculated that cervical ectopic prostatic tissue, vaginal tubulosquamous polyp and similar vulval lesions are derived from eutopic or misplaced paraurethral Skene’s glands which are the female equivalent of prostatic glands in the male (3,7).

Adenoid basal carcinoma is a rare cervical neoplasm accounting for less than 1% of all cervical malignancies (8–12). It is an indolent neoplasm which does not metastasise in pure form; in a review in 2006, Russell and Fadare stated that none of the 66 reported patients with pure adenoid basal carcinoma had experienced tumour recurrence, metastasis or tumour-associated death, irrespective of the modality of treatment (12). Thus alternative terms, such as adenoid basal epithelioma or low-grade adenoid basal tumour, have been proposed to reflect the innocuous behaviour when these neoplasms occur in pure form. Adenoid basal carcinoma is composed of small, bland well-differentiated, rounded nests or sometimes cords of monomorphic basaloid cells with scant cytoplasm. The tumour cell nests may display peripheral palisading, small cysts, cytoplasmic clearing and focal glandular or squamous differentiation; a stromal reaction is usually absent. High-risk human papillomavirus (HPV), most commonly type 16, has been detected in the majority of cases where this has been investigated (8–12). Adenoid basal carcinoma is often associated with a high-grade squamous intraepithelial lesion or with an invasive carcinoma of another type, most commonly squamous cell carcinoma, as seen in our study (8–12). A definitive diagnosis of pure adenoid basal carcinoma therefore requires evaluation of the entire tumour. Adenoid basal carcinoma associated with another invasive carcinoma type should be reported as a mixed carcinoma, with a description of the constituent histotypes and their proportions; the behaviour is dependent on the other tumour type.

The most common antibodies used as tissue markers for prostate cancer are PSA and PSAP. However, these are sometimes expressed at low levels or are absent in prostatic adenocarcinomas and they exhibit lower sensitivity for identifying poorly-differentiated compared to well-differentiated cases; this limits their value in diagnosing prostatic adenocarcinoma in a primary or metastatic site. NKX3.1 is a homeobox gene located on chromosome 8p which shows prostate specific expression, being predominantly localised to the prostatic epithelium (13,14). The gene codes for a transcription factor which functions as a negative regulator of epithelial cell growth in prostatic tissue. Abnormal regulation of this gene is associated with prostate carcinoma progression. It has been demonstrated by Gurel et al that NKX3.1 has high sensitivity and high specificity for metastatic prostatic adenocarcinoma (98.6% sensitivity and 99.7% specificity); in that study, the authors concluded that “NKX3.1 seems to be a highly sensitive and specific tissue marker of metastatic prostatic adenocarcinoma (15)”. Another study by Chuang et al showed that when used as part of a panel of markers, NKX3.1 was highly specific and sensitive in differentiating high-grade prostatic carcinoma from high-grade bladder urothelial carcinoma, a not uncommon diagnostic problem (16).

Two studies have previously examined NKX3.1 immunoreactivity in cervical ectopic prostatic tissue and vaginal tubulosquamous polyp but this is the first study to show expression of this marker in cervical adenoid basal carcinoma (17,18). In the first of these studies, Roma reported two vaginal tubulosquamous polyps and three cases of probable cervical ectopic prostatic tissue, all of which were focally positive for NKX3.1 in the glandular component; PSAP was focally positive in the glandular component in one tubulosquamous polyp and both cases were PSA negative (17). A case of periurethral Skene’s glands was also positive. The author concluded that the findings add support to the postulated origin of vaginal tubulosquamous polyp and cervical ectopic prostatic tissue from eutopic or misplaced Skene glands and that NKX3.1 seems a better marker of these lesions than traditional prostatic markers.

In the second study, Anderson et al studied 13 cases of prostatic tissue involving the vagina (n=12) and ectocervix (n=1) (18). These were mainly in patients with gender dysphoria undergoing long-term testosterone therapy or patients with congenital disorders of sexual development associated with endogenous androgen excess. Immunohistochemically, the prostatic glands stained for NKX3.1 (100% of cases), androgen receptor (100%) and CK7 (92%); PSA was positive in only 69% of cases.

Although we have not systematically studied PSA and PSAP staining in our cases, since these markers are not necessary to diagnose ectopic prostatic tissue and tubulosquamous polyp, it is our experience that these lesions are not always positive with these prostatic markers and when there is immunoreactivity it is often focal and difficult to identify. This was borne out by our results where some cases were negative and some positive (generally focal) when these markers were performed. As such, we agree with Roma that NKX3.1 appears a more robust marker of ectopic prostatic tissue and tubulosquamous polyp than traditional prostatic markers. Moreover, it is a nuclear marker and, as such, is easier to interpret given the problems which sometimes occur in distinguishing between true cytoplasmic staining and a non-specific cytoplasmic blush.

Given the consistent diffuse immunoreactivity of ectopic prostatic tissue, tubulosquamous polyp and adenoid basal carcinoma with NKX3.1, it is likely that this marker will be of value in the diagnosis of these rare entities which, especially the cervical lesions, may have a wide differential diagnosis (19,20). However, this should be proven in future studies which assess NKX3.1 immunoreactivity in a wide range of other cervical glandular lesions. A weakness of our study is that we did not include other benign, premalignant and malignant endocervical glandular lesions. However, in all our cases normal endocervical glands, when present, were negative with NKX3.1.

Three of our 5 cases of adenoid basal carcinoma stained were focally positive with PSAP but negative with PSA. We are not aware of any prior studies having investigated immunoreactivity with traditional prostatic markers or NKX3.1 in cervical adenoid basal carcinoma. The consistent positive staining of this neoplasm with NKX3.1 and positive staining in some cases with PSAP raises the interesting possibility that this indolent carcinoma is somehow associated with ectopic prostatic tissue and could represent the malignant counterpart of this. The coexistence of ectopic prostatic tissue and adenoid basal carcinoma in one of our cases would also be supportive of this. Additionally, these two lesions exhibit some morphological similarities such as the presence of small bland tubular structures. However, adenoid basal carcinoma is usually located at the cervical transformation zone while ectopic prostatic tissue typically is located on the ectocervix. In our case with ectopic prostatic tissue and adenoid basal carcinoma, the two lesions were separate without intermingling, the adenoid basal carcinoma being located at the transformation zone and the ectopic prostatic tissue on the ectocervix. Additionally, as discussed, it is well-established that adenoid basal carcinoma represents an HPV-associated neoplasm and given the differences in morphology (although, as stated, there are also similarities) and the paucity of reports of the coexistence of these two lesions, we feel there is no definitive evidence, at present, of a pathogenic association between them, although this should be explored in further studies. It could be speculated that persistent infection by high-risk HPV of ectopic prostatic tissue predisposes to the development of adenoid basal carcinoma and the common association between this neoplasm and other high-risk HPV-associated carcinomas would be supportive of this.

In summary, we demonstrate consistent nuclear immunoreactivity with the prostatic marker NKX3.1 in cervical ectopic prostatic tissue, vaginal tubulosquamous polyp and cervical adenoid basal carcinoma. While the specificity of this marker should be investigated in future studies, NKX3.1 positivity may be useful in diagnosing these uncommon lesions and confirms that ectopic prostatic tissue and tubulosquamous polyp exhibit “prostatic” differentiation, supporting their derivation from periurethral Skene’s glands. NKX3.1 and PSAP positivity in adenoid basal carcinoma raises the possibility of an association with benign glandular lesions exhibiting prostatic differentiation, although there is limited evidence to support this.

Figure 2.

Low power view of vaginal tubulosquamous polyp composed of well-circumscribed nests of epithelial cells (A). On high power examination, the nests are composed of squamous epithelium with central cyst formation and tubules around the periphery (B). There is diffuse positive staining of the tubular elements with NKX3.1 (C and D).

Figure 3.

Low power of cervical adenoid basal carcinoma showing nests of basaloid cells with small glands infiltrating the stroma without a stromal response (A). At higher power, the nuclear features are bland and eosinophilic material is present within the glandular lumina (B). There is diffuse positive nuclear staining of the glandular component of the adenoid basal carcinoma with NKX3.1 (C and D).