Home About us Editorial board Search Ahead of print Current issue Archives Submit article Instructions Subscribe Contacts Login 
Print this page Email this page Users Online: 257

Ahead of print publication  

Role of cervical cytology in the detection of uterine glandular lesions

 Department of Pathology, Fr Muller Medical College, Mangalore, Karnataka, India

Date of Submission09-Mar-2020
Date of Decision24-Mar-2021
Date of Acceptance14-May-2021

Correspondence Address:
Hilda Fernandes,
Department of Pathology, Fr Muller Medical College, Mangalore - 575 002, Karnataka
Login to access the Email id

Source of Support: None, Conflict of Interest: None

DOI: 10.4103/mjdrdypu.mjdrdypu_68_20


Background: Diagnosis of atypical glandular cells (AGCs) is challenging to the cytopathologist due to the overlapping features of various neoplastic and nonneoplastic lesions. The aim of this study was to assess the utility of conventional Papanicolaou (Pap) smears in detecting uterine glandular lesions. Materials and Methods: Archived records of all histopathologically diagnosed endometrial adenocarcinoma and endocervical adenocarcinoma during the study period were identified. Corresponding Pap smears, if available, were retrieved. In addition, all cytologically reported AGCs and adenocarcinoma during the same period were retrieved. Pap smears of histopathology discordance cases were reviewed. The prevalence of AGC and sensitivity and specificity of Pap smears in detecting glandular lesions were calculated. Results: The prevalence of AGC was 0.65%. There were 11 false negatives and 10 false positives initially, which were reduced to 6 and 5, respectively, after the review. The sensitivity and specificity of Pap smear in detecting AGC was 72.9% and 99.81% initially, which improved marginally to 81.8% and 99.9%, respectively. Conclusions: The prevalence of AGC in Pap smears is low with a moderate sensitivity and good specificity. As the association of clinically significant lesion is high, all cases with AGC diagnosis must be followed up with colposcopic and histopathologic examination of uterus and cervix.

Keywords: Adenocarcinoma, atypical glandular cells, conventional Papanicolaou smear

How to cite this URL:
Fernandes H, Devaraju S, Bhat A, Ramachandra P. Role of cervical cytology in the detection of uterine glandular lesions. Med J DY Patil Vidyapeeth [Epub ahead of print] [cited 2021 Jun 12]. Available from: https://www.mjdrdypv.org/preprintarticle.asp?id=316429

  Introduction Top

Cervical cytological evaluation with Papanicolaou (Pap) smear is the standard screening test for cervical premalignant and malignant lesions. It predominantly detects squamous lesions. Sensitivity in detecting glandular lesions is low due to the sampling technique and rarity of the lesions, with incidence ranging from 0.1% to 2%.[1] The major hurdles to an accurate cytodiagnosis are lesser representation of glandular lesions on  Pap smear More Details and the well-known look-alikes, which often lead to false-positive diagnosis.[2] However, clinically significant lesions have been reported in 18%–43% of patients with atypical glandular cells (AGC) detected on Pap smears.[3] Thus, it is extremely important to correctly identify AGC on Pap smears, followed by immediate colposcopic assessment and fractional curettage.

The present study was undertaken to assess the utility Pap smears (conventional method) in the cytodiagnosis of uterine glandular lesions.

  Materials and Methods Top

This was a retrospective study, conducted over a period of 2 years in the department of pathology at a tertiary care institute. Clearance from institutional ethical committee was obtained (letter no. FMMCIEC/CCM/233/2018 dated April 18, 2018).

The following two-pronged search strategy was adopted [Figure 1]:
Figure 1: Flowchart showing methodology adopted in the study

Click here to view

  1. The histopathology search arm: Archived histopathology records of all cases reported as endometrial adenocarcinoma (EMA) or endocervical adenocarcinoma (ECA) within the study period were identified, and the corresponding Pap smears, if available, were retrieved
  2. The cytopathology search arm: Archived records of all Pap smears reported during the study period were searched for AGC and adenocarcinoma, and Pap smears of such cases were retrieved.

Inclusion criteria

Pap smears of histopathology-reported EMA or ECA were included in the histopathology arm. In the cytopathology arm, Pap smears reported as AGCs subclassified as not otherwise specified (NOS), AGC-favor neoplastic (AGC-FN), adenocarcinoma in situ (AIS), EMA, and ECA were included.

Exclusion criteria

Pap smears reported as negative for intraepithelial neoplasia (NILM), unsatisfactory smears, squamous intraepithelial lesion, and squamous cell carcinoma were excluded.

Analysis of smears

From the cytopathology arm, all the included Pap smear reports were analyzed against their histopathology report to know cytology and histopathology correlation. Similarly, from the histopathology search arm, all the included cases were analyzed against their corresponding Pap smear reports. Discordant cases from both search arms were then reviewed together by two cytopathologists, keeping in mind the histological diagnosis, to see if a revised cytological diagnosis could be offered. The Bethesda system (TBS) 2001 guidelines were followed for Pap smear reporting.

  Results Top

Histopathology search arm

Out of a total 20,443 histopathological specimens processed during the study period, a total of 35 (0.1%) EMA (26) and ECA (9) were identified from the histopathology records [Figure 2]. Pap smears predating the histopathological diagnosis were available in 27 cases. On initial reporting, 16/27 (59.2%) smears were reported as having epithelial cell abnormality (sensitivity -72.9%), 9 as FN, 4 as adenocarcinoma, 2 as atypical squamous cells-cannot exclude high grade, and 1 as adenosquamous carcinoma. Eleven smears were initially reported as NILM. These 11 smears were reviewed in view of the follow-up histopathology diagnosis. After review, 5/11 smears were reported as having epithelial cell abnormality, four as AGC-FN, and one as AGC-endometrial origin. Thus, after the review, 21/27 were reported as having epithelial cell abnormality. The overall sensitivity of Pap smears for the detection of AGC was found to be 81.8% with an overall positivity of 77.77%.
Figure 2: (a) Endometrial carcinoma papillary type (biopsy sample) (H and E, ×400). (b) Endometrioid carcinoma (H and E, ×400). `ometrioid car c inoma with super f icial invas ion of myomet r ium (H and E, ×400). (d) Endocervical carcinoma with cigar-shaped basally located nuclei (H and E, ×400)

Click here to view

Cytopathology search arm

A total of 5457 Pap smears were identified during the study period. Out of these, 36 (0.65%) were reported as AGC/adenocarcinoma. Follow-up histopathology was available in 24 cases. Among these, 14 (58.3%) were positive for malignancy (specificity-99.8%). There were three cases of ECA and eight cases of EMA [Figure 3]. Two cases reported as AGC-FN turned out to be stromal sarcoma and one as carcinosarcoma [Figure 4]. Majority of the patients with a diagnosis of malignancy (all except one) were above 50 years of age with a range of 30–79 years. The most common symptom was postmenopausal bleeding, especially in malignancy. In nonneoplastic conditions, abnormal uterine bleeding was the common symptom, and these women were younger than 50 years of age. Among the ten cases of histologically negative cases, cytological diagnoses remained unchanged in five cases after review. Five cases were downgraded to NILM with reactive glandular cells [Figure 5]. The overall specificity of Pap smears in detecting glandular lesions was 99.9%.
Figure 3: (a) Endometrial adenocarcinoma – tumor cells in papillary pattern (Papanicolaou × 400). (b) Endometrial adenocarcinoma. Tumor cells in sheets and glandular pattern (Papanicolaou × 400). (c) Tumor cells infiltrated by neutrophils (bags of polys) (Papanicolaou × 400). (d) Endocervical adenocarcinoma. Sheets of tumor cells with hyperchromatic, elongated nuclei, feathering and scattered tumor cells in the background (Papanicolaou × 400)

Click here to view
Figure 4: (a) Stromal sarcoma – Tumor cells predominantly in singles. Reported as atypical glandular cell-favor neoplastic in cytology (Papanicolaou × 400). (b) Histopathology of endometrial stromal sarcoma (H and E, ×200)

Click here to view
Figure 5: (a) Exfoliated endometrial cells (exodus ball) (Papanicolaou × 200). (b) Atypical repair – sheets of cells with enlarged nuclei, coarsec hromatin (Papanicolaou × 400). Reported as atypical glandular cell-not otherwise specified. (c) Endocervical polyp – hyperchromatic nuclei with peripheral fanning (Papanicolaou × 400). Reported as atypical glandular cell-not otherwise specified (endocervical). (d) Abraded lower uterine segment – irregular sheets and tubules of endometrial cells (Papanicolaou × 200), reported as atypical glandular cell-not otherwise specified (endometrial)

Click here to view

  Discussion Top

Glandular lesions of the female genital tract (FGT) are quite uncommon compared to squamous lesions. Their cytological diagnosis is difficult because of their architectural and cytological complexity, as well as low sensitivity due to sampling techniques.[4] The diagnostic category of AGC was introduced in the 1988 Bethesda system.[5] The frequency of reporting glandular lesions is witnessing an upward trend in recent years.[6],[7] In our study, the overall incidence of AGC was 0.65%. This is similar to the other studies in the literature.[3],[7] The age of patients with glandular lesions ranged from 30 to 79 years. Sixty-one percent (22 cases) of the women were postmenopausal. Postmenopausal bleeding was the common symptom (44.4%), followed by abnormal uterine bleeding (19.4%). These findings are similar to the findings described by other authors.[8],[9],[10],[11] Histopathological follow-up was available in 24 patients. Among these 24 patients, a clinically significant lesion was confirmed in 14 patients (58.3%). Mood et al. have reported similar percentage (55.3%) of clinically significant lesions,[12] whereas Kim et al.,[6] Krane et al.,[13] and Boyraz et al.[9] have reported lower percentage of clinically significant histologically confirmed lesions. When the association with significant pathology was evaluated according to different AGC subgroups, a diagnosis of AGC-FN was the most serious one that was associated with invasive disease (all cases). Similar findings were observed by others.[9],[11]

The most common origin of significant pathology in the present study was adenocarcinoma endometrium (8/14), followed by cervix (3/14). According to one of the largest published series, EMA was the most commonly diagnosed malignancy with AGC.[14] The cervical smears will yield a positive diagnosis in about 25% of cases of endometrial carcinomas.[15] The higher incidence in our study may be attributed to the lower incidence of adenocarcinoma cervix compared to endometrial neoplasia.[6],[16] This finding is also affirmed by the histopathology search arm conducted in our study, which showed 9/35 cases of cervical adenocarcinoma.

Although well-differentiated endometrioid carcinomas have a reasonably characteristic cytologic presentation, the precise histologic type of endometrial carcinoma can rarely be established in cytologic material. High-grade endometrioid carcinomas, their variants, and serous-papillary carcinomas shed similar cells. When EMAs shed papillary cell clusters, the differential diagnosis must comprise adenocarcinomas of the  Fallopian tube More Details and ovary and adenocarcinomas of other origins metastatic to the FGT. If only single, large cancer cells are present in the cytologic sample, the differential diagnosis should include other cancers, such as a poorly differentiated squamous (epidermoid) carcinoma and other poorly differentiated primary or metastatic tumors.[15]

We had five false-positive cases even after review. False positives were seen only in the category of AGC-NOS. Majority of the cases reported as AGC-NOS had endocervical cells showing reactive/reparative changes, which are difficult to differentiate from malignancy. On histopathology, these showed papillary endocervicitis, endocervical polyp, and microglandular adenosis. This highlights the limitation of the cervical cytology for the precise diagnosis of AGC. Similar results have been reported by other authors.[8],[17],[18] The high rates of false positivity may be due to overinterpretation of the reactive atypia of glandular cells. In Pap smears, glandular cells can be from normal endocervical cells, reactive and reparative endocervical cells, cervical AIS, and adenocarcinoma cervix. Endometrial cells may be from menstrual endometrium, hyperplasia, polyps, and adenocarcinoma. Tubal metaplasia, intra uterine contraceptive device changes, or direct sampling from lower uterine segment may pose diagnostic challenge. Glandular cells from extrauterine malignancies such as tubal, ovarian, and breast origin may be seen.[4] Therefore, it is important to carefully analyze the smears using stringent diagnostic criteria of AGC. A close attention to the architecture, nuclear membrane, chromatin pattern, and dissociated atypical cells will help in distinguishing between neoplastic and benign lesions.[3]

We had two cases of stromal sarcoma and one case of carcinosarcoma in our study, which were reported as AGC-FN in cytology. These smears showed singly dispersed naked nuclei. Characteristic stroma or comet cells were not seen. The background was clean. Finding tumor cells of these lesions in Pap smear is rare.[19] Policarpio et al. observed scanty tumor cells in one out of 12 smears of low-grade stromal sarcomas.[20] On Pap test, differential diagnosis of low-grade endometrial stromal sarcoma (ESS) includes normally shed endometrial cells, endometriosis and submucosal leiomyoma, and carcinosarcoma. For high-grade ESS, a differential diagnosis of lymphoma and poorly differentiated carcinoma has to be considered. In our single case of carcinosarcoma, the sarcoma component was not observed even after extensive search. Snyder et al. reported a 60% sensitivity for detecting carcinosarcoma and inferred that it is an adverse prognosticator.[21]

There were 6/27 cases of false negatives in our study. Majority of these cases were from endometrial carcinomas. These smears did not show any atypical cells even after extensive review. This may be due to the lesser exfoliation of endometrial cancer cells into the cervical canal. Studies have shown that the use of liquid-based cytology (LBC) techniques improves the detection rate of glandular cells, especially for endometrial cancer.[8],[11],[22] This is mainly because of cell concentration and better preservation of cellular morphology in LBC samples, in contrast to the conventional smears.

The overall positivity (after review) in our study was 77.7%. The overall positivity of diagnosing glandular malignancy in the Pap smear varies for cervical and endometrial lesions. Geldenhuys and Murray reported that 80% of the Pap smears with ECA and 22% with EMA had positive findings within a year of the histologic diagnoses, with 51% overall positivity.[23] Bansal et al. reported an overall positivity of 41.8%.[8] Moreira et al. observed a sensitivity ranging from 55.8% to 73.1% and a specificity ranging from 66.1% to 87.1% in the interobserver variation study.[17] In our study, a sensitivity of 72.9% and specificity of 99.8 % was observed which showed slight improvement on review. Use of LBC technique, combining with human papillomavirus (HPV) detection, may help minimize the limitations of conventional Pap smear in diagnosing AGC, provided endometrial cancer is ruled out. Studies have shown that about 24%–45% of AGC diagnosed on cytology test positive for high-risk HPV DNA.[24],[25]

  Conclusions Top

The prevalence of AGC in Pap smears is low with a low-to-moderate sensitivity and specificity. Glandular lesions are diagnostically and therapeutically challenging because of their rarity, absence of colposcopic findings, irregular shedding, small size, higher location, and broader differential diagnoses. The diagnosis of AGC in cytology is associated with significant pathology in a considerable proportion of patients. Therefore, these patients should be evaluated vigilantly with histological workup, especially if they are older than 50 years and postmenopausal.

Financial support and sponsorship


Conflicts of interest

There are no conflicts of interest.

  References Top

Jeng CJ, Liang HS, Wang TY, Shen J, Yang YC, Tzeng CR. Cytologic and histologic review of atypical glandular cells (AGC) detected during cervical cytology screening. Int J Gynecol Cancer 2003;13:518-21.  Back to cited text no. 1
Johnson JE, Rahemutulla A. Endocervical neoplasia and its mimics in Thin Prep Pap tests. A descriptive study. Acta Cytol 1999;43:369-75.  Back to cited text no. 2
Ajith D, Gavas S, Joseph S, Rekhi B, Deodhar K, Kane S. Identification of atypical glandular cells in Pap smear. Is it a hit and miss scenario? Acta Cytol 2013;57:45-53.  Back to cited text no. 3
Covell JL, Wilbur DC, Guidos B, Lee KH, Chhieng DC, Mody DR. Epithelial abnormalities: Glandular. In: Solomon D, Nayar R, editors. The Bethesda System for Reporting Cervical Cytology. 3rd ed.. New York: Springer; 2015. p. 204-6.  Back to cited text no. 4
The 1988 Bethesda System for reporting cervical/vaginal cytological diagnoses. National Cancer Institute Workshop. JAMA 1989;262:931-4.  Back to cited text no. 5
Kim SS, Suh DS, Kim KH, Yoon MS, Choi KU. Clinicopathological significance of atypical glandular cells on Pap smear. Obstet Gynecol Sci 2013;56:76-83.  Back to cited text no. 6
Smith HO, Tiffany MF, Qualls CR, Key CR. The rising incidence of adenocarcinomas relative to squamous cell carcinoma of the uterine cervix in United States, a 24 –year population study. Gynecol Oncol 2000;78:97-105.  Back to cited text no. 7
Bansal B, Gupta P, Gupta N, Rajwanshi A, Suri V. Detecting uterine glandular lesions: Role of cervical cytology. Cytojournal 2016;13:3.  Back to cited text no. 8
[PUBMED]  [Full text]  
Boyraz G, Basaran D, Salman MC, Ibrahimov A, Onder S, Akman O, et al. Histological follow-up in patients with atypical glandular cells on Pap smears. J Cytol 2017;34:203-7.  Back to cited text no. 9
[PUBMED]  [Full text]  
Suresh PK, Kini H, Minal J, Dhavalpure N, Basavaiah SH, Adiga DS, et al. The significance of glandular cells on conventional cervicovaginal smears: Experience from a tertiary care hospital in Coastal India. Acta Cytol 2017;61:199-206.  Back to cited text no. 10
Cangiarella JF, Chhieng DC. Atypical glandular cells – An update. Diagn Cytopathol 2003;29:271-9.  Back to cited text no. 11
Mood NI, Eftekhar Z, Haratian A, Saeedi L, Mogaddam PR, Yarandi F. A cytohistologic study of atypical glandular cells in conventional Pap smears in a large, high risk U.S West coast minority population. Acta Cytol 2009;53:153-9.  Back to cited text no. 12
Krane JF, Lee KR, Sun D, Yuan L, Crum CP. Atypical glandular cells of undetermined significance. Outcome predictions based on human papillomavirus testing. Am J Clin Pathol 2004;121:87-92.  Back to cited text no. 13
Schnatz PF, Guile M, O'Sullivan DM, Sorosky JI. Clinical significance of atypical glandular cells in cervical cytology. Obstet Gynecol 2006;107:701-8.  Back to cited text no. 14
Koss LG. Proliferative disorders and carcinoma of the endometrium. In: Koss LG, Melamed MR, editors. Koss' Diagnostic Cytology and its Histopathologic Basis. 5th ed.. Philadelphia: Lippincott Williams & Wilkins; 2006. p. 442.  Back to cited text no. 15
Chhieng DC, Elgert P, Cohen JM, Cangiarella JF. Clinical significance of atypical glandular cells of undetermined significance in postmenopausal women. Cancer 2001;93:1-7.  Back to cited text no. 16
Moreira MA, Longatto Filho A, Castelo A, de Barros MR, Silva AP, Thomann P, et al. How accurate is cytological diagnosis of cervical glandular lesions? Diagn Cytopathol 2008;36:270-4.  Back to cited text no. 17
Nanda K, McCrory DC, Myers ER, Bastian LA, Hasselblad V, Hickey JD, et al. Accuracy of the Papanicolaou test in screening for and follow-up of cervical cytologic abnormalities: A systematic review. Ann Intern Med 2000;132:810-9.  Back to cited text no. 18
Bardarov S, Khachaturov V, Kirtesh P, Jimenez E. Case report of undifferentiated endometrial sarcoma in association with osteoclast-like giant cells. Patholog Res Int 2011;2011. doi: 10.4061/2011/629840.  Back to cited text no. 19
Policarpio ML, Cathro HP, Kerr SE, Stelow EB. Cytomorphological features of low-grade endometrial stromal sarcoma. Am J Cl Pathol 2007;128:265-71.  Back to cited text no. 20
Snyder MJ, Robboy SJ, Vollmer RT, Dodd LG. An abnormal cervicovaginal cytology smear in uterine carcinosarcoma is an adverse prognostic sign: Analysis of 25 cases. Am J Clin Pathol 2004;122:434-9.  Back to cited text no. 21
Wood MD, Horst JA, Bibbo M. Weeding atypical glandular cell look-alikes from the true atypical lesions in liquid-based Pap tests: A review. Diagn Cytopathol 2007;35:12-7.  Back to cited text no. 22
Geldenhuys L, Murray ML. Sensitivity and specificity of the Pap smear for glandular lesions of the cervix and endometrium. Acta Cytol 2007;51:47-50.  Back to cited text no. 23
Wright TC. Management of women with high grade squamous intraepithelial lesion and atypical glandular cell cervical cytology. In: Monsonégo J, editor. Emerging Issues on HPV Infections: From Science to Practice. Basel: Karger; 2006. p. 140-6.  Back to cited text no. 24
Miller RA, Mody DR, Tams KC, Thrall MJ. Glandular lesions of the cervix in clinical practice: A cytology, histology, and human papillomavirus correlation study from 2 institutions. Arch Pathol Lab Med 2015;139:1431-6.  Back to cited text no. 25


  [Figure 1], [Figure 2], [Figure 3], [Figure 4], [Figure 5]


     Search Pubmed for
    -  Fernandes H
    -  Devaraju S
    -  Bhat A
    -  Ramachandra P
    Access Statistics
    Email Alert *
    Add to My List *
* Registration required (free)  

  In this article
   Materials and Me...
   Article Figures

 Article Access Statistics
    PDF Downloaded1    

Recommend this journal