G Model PRP-51594; No. of Pages 4
ARTICLE IN PRESS Pathology – Research and Practice xxx (2016) xxx–xxx
Contents lists available at ScienceDirect
Pathology – Research and Practice journal homepage: www.elsevier.com/locate/prp
Original article
Cervical polyps: Is histologic evaluation necessary? Rebecca A. Levy ∗ , Asangi R. Kumarapeli, Horace J. Spencer, Charles M. Quick University of Arkansas for Medical Sciences, Department of Pathology, Slot 517, 4301 West Markham St. Little Rock, AR 72205, United States
a r t i c l e
i n f o
Article history: Received 27 April 2016 Received in revised form 1 June 2016 Accepted 22 June 2016 Keywords: Cervical polyps Cervical dysplasia Cervical malignancy Mullerian adenosarcoma
a b s t r a c t Objective: The aim of this study was to examine a series of clinically identified cervical polyps and determine the incidence of significant histologic and concurrent cytologic findings. Methods: Consecutive cervical polyps from January 2000 through September 2012 were retrieved from the hospital laboratory information system. Histologic evaluation of these polyps was performed, followed by a chart review of clinical findings and correlation with the immediately prior or concurrently collected cervical Papanicolaou (Pap) test results, when available. Results: A total of 369 cervical polyps were identified and reviewed. The patient ages ranged from 18 to 87 years (mean 46.5 years). Eight polyps demonstrated squamous dysplasia (6 Cervical Intraepithelial Neoplasia/CIN I, and 2 CIN II/III), while 6 had malignant or atypical/potentially malignant features (2 adenosarcoma, 2 atypical polyps concerning for Mullerian adenosarcoma, 1 endometrioid endometrial adenocarcinoma and 1 adenocarcinoma in-situ). An increased incidence of atypical squamous cells of undetermined significance (ASCUS) and atypical glandular cells not otherwise specified (AGC NOS) Pap diagnoses (12.7% and 6.1%, respectively) was found in women with benign polyps on biopsy. Discussion: We demonstrated a higher rate of clinically significant histologic findings in cervical polyps (14 of 369 cases, 3.7%) compared to previously reported studies. The increase in ASCUS and AGC Pap results was most likely related to reactive and inflammatory changes present in benign polyps. Our results suggest that removal of all cervical polyps with subsequent histologic review is warranted. © 2016 Elsevier GmbH. All rights reserved.
1. Introduction Cervical polyps arise within the endocervical canal and are reportedly present in 2–5% of women of reproductive age [1]. Cervical polyps are composed of papillary proliferations of epithelial tissue around a fibrovascular stromal core which may have glandular or squamous epithelium. Cervical polyps are frequently asymptomatic and can be discovered upon routine gynecologic examination whereas symptomatic polyps commonly present with intermenstrual, postcoital and/or postmenopausal bleeding [2]. Large clinical studies have shown that cervical polyps are usually benign, with a prevalence of malignancy in only 0.1% of cases [3,4]. The underlying pathophysiology of cervical polyps is uncertain;
Abbreviations: Pap, Papanicolaou test; hr-HPV, high-risk Human Papilloma Virus; ASCUS, atypical squamous cells of undetermined significance; AGC NOS, atypical glandular cells not otherwise specified; ASC-H, atypical squamous cells: cannot exclude high-grade squamous intraepithelial lesion; LSIL, low-grade squamous intraepithelial lesion; HSIL, high-grade squamous intraepithelial lesion. ∗ Corresponding author. E-mail addresses:
[email protected] (R.A. Levy),
[email protected] (A.R. Kumarapeli),
[email protected] (H.J. Spencer),
[email protected] (C.M. Quick).
possible causes include chronic inflammation, hormonal stimulation, and/or congestion of cervical blood vessels [5]. Recently published studies state that histologic examination of all cervical polyps may not be necessary because the likelihood of finding a malignant lesion is so low, and concluded that in asymptomatic polyps there was no indication for polypectomy and histologic examination if the concurrent Pap test is normal [6–8]. Furthermore, it has been reported that the most common malignancy identified in cervical polyps is endometrial in origin, and that primary cervical/endocervical malignancy in a cervical polyp is very rare [4]. A recent British study evaluated the cost efficacy of performing polypectomies on symptomatic cervical polyps, or polyps with concurrent atypical Pap test results, as opposed to the removal of all cervical polyps7 . The majority (67%) of polyps in their study were asymptomatic, and no cases of malignancy were found in 1366 cervical polyps. Given the current health care environment, there is concern that the unnecessary removal of polyps can have a significant effect on health care resources. This study concluded that significant decrease in health care costs could be attained if asymptomatic polyps were not removed [7].
http://dx.doi.org/10.1016/j.prp.2016.06.010 0344-0338/© 2016 Elsevier GmbH. All rights reserved.
Please cite this article in press as: R.A. Levy, et al., Cervical polyps: Is histologic evaluation necessary? Pathol. – Res. Pract (2016), http://dx.doi.org/10.1016/j.prp.2016.06.010
G Model PRP-51594; No. of Pages 4
ARTICLE IN PRESS R.A. Levy et al. / Pathology – Research and Practice xxx (2016) xxx–xxx
2
Conversely, there are experts who recommend routine removal of all cervical polyps [4,9,10]. The studies that support this approach argue that removal of all cervical polyps is helpful to rule out malignancy or explain atypical Pap test results. Two studies have reported higher rates of atypia and dysplasia in endocervical polyps in premenopausal patients and an increased frequency of malignancy in cervical polyps in post-menopausal patients [9,11]. Fauth et al. [9] reported an incidence of dysplasia and malignancy in 1.4% of 4402 cervical polyps examined, and Schnatz et al. [4] evaluated 2458 polyps and reported rates of cervical malignancy, dysplasia and atypia at 0.1%, 0.7%, and 1.9%, respectively in women younger than 50 years (premenopausal women). Cervical polyps may also be the presenting finding of carcinoma from elsewhere in the gynecologic tract. A Turkish study of 4063 patients found 3 cases of metastatic endometrial adenocarcinoma in cervical polyps of postmenopausal women [11]. They concluded that endometrial pathology may be associated with endocervical polyps in the postmenopausal group. Studies have found increased incidence of endometrial malignancies and hyperplasia in patients with postmenopausal bleeding [12,13]. Benign polyps have been associated with squamous metaplasia, microglandular hyperplasia, inflammation, erosion, Arias-Stella reaction, and endometriosis [10]. Morphologic findings of reactive changes may be interpreted as atypical findings on cytologic evaluation. Prior studies have shown that a significant number of AGC Pap test results are related to benign endocervical polyps; Cheng [14] reported 21% of AGC cases had histologic findings of benign polyps and Sorosky [3] reported that 71% of AGC pap tests corresponded to reactive changes, benign polyps and normal endocervical histology.
2. Methods All consecutive endocervical polyps evaluated at the University of Arkansas for Medical Sciences (UAMS) between January 2000 and September 2012 were retrieved from electronic medical records. The search terms “cervix”, “endocervix”, “cervical polyp”, “endocervical polyp, and ‘atypical polyp’ were used to identify cases within the clinical information and final diagnosis fields of the pathology reports. The laboratory information system was also utilized to obtain clinical information that included patient demographics, clinical presentation, type of procedure and histologic diagnosis. Results of cervical Pap tests performed at the time of cervical biopsy or within one month before or after the procedure were collected with high-risk Human Papilloma Virus (hr-HPV) DNA results when available. This study was approved by the UAMS Institutional Review Board. Specimens with scant tissue that did not meet the histologic criteria for a cervical polyp were excluded. All cases were reviewed by a pathologist (RAL) and atypical findings were confirmed by a second pathologist (CMQ). The polyps were evaluated for the following histomorphologic features: inflammation, thick walled vessels, reactive epithelium, squamous metaplasia, tubal/endometrial metaplasia, microglandular hyperplasia, endometriosis, atypical stromal cells, stromal mitoses, granulation/ulceration, cervical intraepithelial neoplasia, and evidence of malignancy. We analyzed the data under two age groups; premenopausal (age <50 years) and post-menopausal (≥50 years). Categorization of menopausal status of patients by age is not ideal, however, it has been used in previously reported studies [2,4,6]. Histologic findings of the biopsies were correlated with Pap results and hr-HPV DNA results, when available. Microsoft Excel was used to analyze the data. Contingency tables were analyzed with Fisher’s exact test and p < 0.05 was considered as significant.
Table 1 Breakdown of age ranges of 369 women with cervical polyps. Age (years)
N
%
Total 18–20 21–29 30–39 40–49 50–59 60+ Unknown
369 3 33 73 119 96 43 2
100 0.8 8.9 19.8 32.2 26.0 11.7 0.5
Table 2 Incidence of reactive histologic findings in benign cervical polyps. Benign and Reactive Findings
Incidence
Inflammation Thick walled vessels Reactive epithelium Squamous metaplasia Tubal/endometrial metaplasia Multiglandular hyperplasia Endometriosis Atypical stroma Stromal mitoses Granulation/Ulceration
356 (96.5%) 361 (97.8%) 336 (91.1%) 145 (39.3%) 79 (21.4%) 67 (18.2%) 18 (4.9%) 29 (7.9%) 5 (1.3%) 53 (14.1%)
3. Results A total of 369 cervical polyps were identified and reviewed. Patient age ranged from 18 to 86 (mean age 46.5 years) (see Table 1). Three hundred fifty-six of these polyps (96.47%) showed reactive findings, and over 90% of them contained the characteristic findings of inflammation, thick-walled vessels, and reactive epithelium. Squamous, tubal and endometrial metaplasia were common findings in the benign polyps as well (see Table 2). Seventyseven women with cervical polyps presented with vaginal bleeding (20.87%) while the remaining were either asymptomatic or the reason for presentation was not documented. Pap test results were available in 228 patients with polyps. Sixtyfour of them had abnormal cytologic findings (29 ASCUS, 14 cases of AGC NOS, 3 ASC-H (Atypical squamous cells: cannot exclude high-grade squamous intraepithelial lesion), 15 LSIL (low-grade squamous intraepithelial lesion) and 3 HSIL (high-grade squamous intraepithelial lesion)), yet benign histologic features on corresponding polyps. In our study population, hr-HPV DNA molecular tests were available on 8 patients with Pap tests showing atypical squamous cells of uncertain significance (ASCUS). Six of the 8 patients had positive hr-HPV results. Of the remaining Pap tests with ASCUS, either hr-HPV DNA testing was not performed in house or the Pap test was performed prior to 2003 when hr-HPV DNA tests were not routinely use. Regardless of the HPV status however, all patients with ASCUS on Pap tests showed histologically benign cervical polyps. We identified 14 cases (3.7%) with clinically significant histologic findings consisting of squamous dysplasia, atypical polyps, and malignancies in polyps (see Table 3). Eight polyps showed squamous dysplasia including 6 cases of LSIL (CIN I) and 2 cases of HSIL (CIN II–III) on the polyp surface; 2 polyps had adenosarcoma, 2 polyps demonstrated atypical features concerning for a Mullerian adenosarcoma, 1 polyp was involved by endometrioid endometrial adenocarcinoma and 1 polyp displayed adenocarcinoma in-situ (see Fig. 1). The 8 polyps with dysplasia accounted for 2.17% of all polyps examined and the 6 polyps with atypical and malignant findings accounted for 1.63%. Within the premenopausal age group, 12 of 229 patients (5.24%) were found to have dysplastic or malignant results while in the postmenopausal group they accounted for 1.43% (2 of 140 patients) (Fisher’s Exact Test p-value = 0.0903).
Please cite this article in press as: R.A. Levy, et al., Cervical polyps: Is histologic evaluation necessary? Pathol. – Res. Pract (2016), http://dx.doi.org/10.1016/j.prp.2016.06.010
G Model PRP-51594; No. of Pages 4
ARTICLE IN PRESS R.A. Levy et al. / Pathology – Research and Practice xxx (2016) xxx–xxx
3
Table 3 Fourteen cervical polyps with cervical dysplasia and malignancies. None of these cases had HPV testing. Histologic Finding
Patient age
Polyp Size
Presentation
Cytologic findings
LSIL (CIN I) LSIL (CIN I) LSIL (CIN I) LSIL (CIN I) LSIL (CIN I) LSIL (CIN I) HSIL (CIN II–III) HSIL (CIN II–III) Adenosarcoma Adenosarcoma Atypical polyp Atypical polyp Endometrioid adenocarcinoma
22 41 55 35 43 31 44 35 33 74 48 23 33
0.3 cm 0.4 cm 1.2 cm 0.8 cm 0.7 cm 1.2 cm 2.0 cm 1.8 cm 2.0 cm 2.5 cm 2.0 cm 3.9 cm 3.9 cm
Colposcopy Colposcopy Prior polyp Colposcopy Cervical lesion Not available Uterine polyp Not available Cervical lesion Prior polyp Prolapsed fibroid Cervical polyp Cervical lesion
LSIL LSIL LSIL LSIL NILM HSIL NILM No Pap test No Pap test No Pap test No Pap test NILM Adenocarcinoma
Fig. 1. Malignancy identified within endocervical polyps. A: Adenosarcoma (low power, 4X, H&E stain) with B: Intact cambium (high power, 20X, H&E stain). C. Endometrial adenocarcinoma (low power, 4X, H&E stain) with D. Endometrial lining (high power, 20X, H&E stain).
Seven out of 8 patients who demonstrated squamous dysplasia on the cervical polyps had Pap tests performed. One of them was negative for intraepithelial lesion or malignancy (+NILM), 5 showed LSIL and one had HSIL. Of the six cases with malignant findings on the polyps, only two patients had corresponding Pap tests that showed adenocarcinoma in one and NILM in the other. 4. Discussion Overall, our results indicated a higher incidence of dysplastic (2.17%) and atypical/malignant (1.63%) findings in cervical polyps than reported in prior studies [1–4,6–11]. Two of the polyps in the atypical/malignant category were suspicious for a Mullerian adenosarcoma. Both demonstrated altered architecture and periglandular stromal cuffing and one polyp had atypical stroma
and the other displayed rare (<2/HPF) atypical mitoses. Howitt et al. defined atypical polyps as those having some, but not all features of Mullerian adenosarcoma, which included phyllodes-like architecture, periglandular stromal alteration, atypical stromal cells and mitoses ≥2/10 HPF [14]. Howitt provided histologic and clinical follow-up in the majority of patients, 86% and 97% respectively, and the results showed a benign clinical course in patients with atypical polyps [14]. Prior studies have shown that premenopausal patients have a higher rate of dysplasia residing in cervical polyps than postmenopausal patients [4,9,11]. These results were supported by our study in which 7 out of 8 patients with polyps containing dysplasia were found in premenopausal women. There were three times more abnormal findings in younger patients than in older patients. Therefore, it is important that all polyps undergo histopathologic
Please cite this article in press as: R.A. Levy, et al., Cervical polyps: Is histologic evaluation necessary? Pathol. – Res. Pract (2016), http://dx.doi.org/10.1016/j.prp.2016.06.010
G Model PRP-51594; No. of Pages 4
ARTICLE IN PRESS R.A. Levy et al. / Pathology – Research and Practice xxx (2016) xxx–xxx
4
evaluation so that early detection of dysplastic findings in the premenopausal population are not missed. Our results may be biased due to poor access to health care and late presentation of patients in our geographical region. Recent state and national data indicate that Arkansas has an increased rate of cervical cancer and increased mortality rate in cervical cancer in comparison to the national rates (data obtained between 1997 and 2005)[15,16]. Within this time frame, Arkansas’ incidence of cervical carcinoma was 2–4% higher than the national average in 5 out of 8 years (1997, 1999, 2001, 2004 and 2005), and in 2002 Arkansas’ incidence of mortality due to cervical carcinoma was almost twice the national rate [15,16]. The patient population in this study included women who often did not have concurrent or prior pap tests (38%, 141/369 patients with no Pap tests), which may indicate lack of access to routine health care and thus an increased incidence of cervical polyp dysplasia and malignancy at presentation. Abnormal Pap test results can be associated with regenerative/inflammatory processes as well as malignancy. In our review, Pap test results were available in 62% (228/369) of the total cases. Our results demonstrated ASCUS and AGC cytology rates of 12.7% and 6.1%, respectively, in women with benign polyps. The majority of patients in our study with abnormal prior or concurrent pap test results had benign endocervical polyps without evidence of dysplasia or malignancy. Our results supported previous observations in that benign polyps often contained reactive changes, which could have led to atypical cells on Pap tests [12,17,18]. Therefore, if a woman has a Pap test indicative of ASCUS or AGC in the setting of a cervical polyp, then polypectomy, cervical biopsy or endocervical curettage should be performed as the polyp may explain the atypical cells seen on the Pap test. It is difficult to determine whether concurrent Pap test could reliably distinguish malignant polyps from reactive ones. In our study, four of six patients with malignant polyps had no corresponding Pap tests. Of the remaining 2 patients, 1 had a negative pap test, and 1 was positive for adenocarcinoma. Hence, no conclusions can be drawn as to the utility of abnormal Pap tests in identifying malignancy in cervical polyps. Our analysis of hr-HPV DNA results was not illuminating due to the small number of cases available to compare with the cervical polyps. Reflex HPV testing was performed on a small fraction of abnormal Pap tests (particularly ASCUS results) and the lack of results for all ASCUS pap smears in our study was due to multiple factors. Reasons include the fact that not all Pap tests that corresponded with the cervical biopsies were performed at our institution, inadequate patient follow-up and inadequate samples for molecular testing. Furthermore, our study population included patients from year 2000 at a time when the molecular testing was not routinely used in the USA (hr-HPV DNA testing was introduced and implemented in 2003). As a result, we were unable to provide a meaningful correlation between the limited number of hr-HPV DNA test results available and the diagnosis of cervical polyps in the presence of atypical Pap results. 5. Conclusion
nancy in a large series of consecutive cervical polypectomies, and to compare these findings with available Pap test results. Our results demonstrate that dysplastic or malignant conditions can often be identified in clinically innocuous cervical polyps, thus necessitating their removal for evaluation. Furthermore, removal and evaluation of all cervical polyps may help explain abnormal Pap test findings, which in turn, may ease consternation in patients and gynecologists. Therefore, we stress the importance of pathologic evaluation of all cervical polyps, regardless of symptoms or abnormal Pap tests. Declaration of interests The authors have no conflicts of interest to declare. References [1] H.K. Farrar, R. Nedoss, Benign tumors of the uterine cervix, Am. J. Obstet. Gyenecol. 81 (1961) 124. [2] S.A. Tirlapur, A. Adeyemo, N. O’Gorman, D. Selo-Ojeme, Clinico-pathological study of cervical polyps, Arch. Gynecol. Obstet. 282 (2010) 535–538. [3] C.E. Berzolla, P.F. Schnatz, D.M. O’sullivan, R. Bansul, S. Mandavilli, Sorosky JI. dysplasia and malignancy in endocervical polyps, J. Womens Health 16 (9) (2007) 1317–1321. [4] P.F. Schnatz, S. Ricci, D.M. O’sullivan, Cervical polyps in postmenopausal women: is there a difference in risk? Menopause 16 (3) (2009) 524–528. [5] I. Stamatellos, P. Stamatopoulos, J. Bontis, The role of hysteroscopy in the current management of the cervical polyps, Arch. Gynecol. Obstet. 276 (4) (2007) 299–303. [6] O. Goldshmid, E. Schejter, D. Kugler, J. Menczer, Is removal of asymptomatic cervical polyps necessary? histologic findings in asymptomatic israeli jewish women, J. Low. Genit. Tract Dis. 15 (4) (2011) 259–262. [7] I.Z. MacKenzie, C. Naish, C.M.P. Rees, S. Manek, Why remove all cervical polyps and examine them histologically? BJOG 116 (2009) 1127–1129. [8] M.T. Younis, S. Iram, B. Anwar, Women with asymptomatic cervical polyps may not need to see a gynaecologist or have them removed: an observational retrospective study of 1126 cases, Eur. J. Obstet. Gynecol. Reprod. Biol. 150 (2010) 190–194. [9] C. Fauth, A. Franko, Q. Duan, S. Wood, M.A. Duggan, Clinicopathological determinants of vaginal and premalignant-malignant cervico-vaginal polyps of the lower female genital tract, J. Low. Genit. Tract Dis. 15 (3) (2011) 210–218. [10] M.E. Long, D.S. Dwarica, T.M. Kastner, M.M. Gallenberg, P.D.M. Chantigian, M.L. Marnach, A.L. Weaver, P.M. Casey, Comparison of dysplastic and benign endocervical polyps, J. Low. Genit. Tract Dis. 17 (2013) 142–146. [11] E.E. Buyukbayrak, A.Y.K. Karsidag, B. Kars, et al., Cervical polyps: evaluation of routine removal and need for accompanying D&C, Arch. Gynecol. Obstet. 283 (2011) 581–584. [12] J. Kaur, P. Dey, S.C. Saha, et al., Cervical cytology in patients with postmenopausal bleeding, Diagn. Cytopathol. 38 (7) (2009) 496–498. [13] M.U. Tariq, R. Idrees, A. Raheem, N. Kayani, Spectrum of histopathological findings in postmenopausal bleeding, Coll. Phys. Surg. Pak 25 (11) (2015) 794–797. [14] B.E. Howitt, B.J. Quade, M.R. Nucci, Uterine polyps with features overlapping with those of mullerian adenosarcoma a clinicopathologic analysis of 29 cases emphasizing their likely benign nature, Am. J. Surg. Pathol. 39 (1) (2015 Jan) 116–126. [15] Arkansas Cancer Registry, Cancer statistics reports for the state of Arkansas. Available at http://cancer-rates.info/ar, (accessed 08.05.13). [16] Surveillance, Epidemiology, and End Results Program, Cancer statistics reports for the state of Arkansas. Available at www.seer.cancer.gov, (accessed 08.05.13). [17] P.F. Schnatz, M. Guile, D.M. O’sullivan, J.I. Sorosky, Clinical significance of atypical glandular cells on cervical cytology, Obstet. Gynecol. 107 (3) (2006 Mar) 701–708. [18] R.F. Cheng, E. Hernandez, L.L. Anderson, P.B. Heller, R. Shank, Clinical significance of a cytologic diagnosis of atypical glandular cells of undetermined significance, J. Reprod. Med. 44 (11) (1999) 922–928.
The purpose of this study was to evaluate the histologic features of cervical polyps, assess the prevalence of dysplasia and malig-
Please cite this article in press as: R.A. Levy, et al., Cervical polyps: Is histologic evaluation necessary? Pathol. – Res. Pract (2016), http://dx.doi.org/10.1016/j.prp.2016.06.010