Can concurrent high-risk endometrial carcinoma occur with atypical endometrial hyperplasia?

(1)

Contents lists available atScienceDirect

International Journal of Surgery

journal homepage:www.elsevier.com/locate/ijsu

Original Research

Can concurrent high-risk endometrial carcinoma occur with atypical

endometrial hyperplasia?

Baki Erdem

a,∗

, Osman A

şıcıoğlu

a

, Niyazi Alper Seyhan

a

, Nuri Peker

b

, Volkan Ülker

a

,

Özgür Akbay

ır

a

a_{Department of Gynecologic Oncology, Kanuni Sultan Süleyman Training and Research Hospital, Istanbul, Turkey} b_{Department of Obstetrics and Gynecology, Istinye University Bahçeşehir Liv Hospital, Istanbul, Turkey}

A R T I C L E I N F O

Keywords:

Atypical endometrial hyperplasia Endometrial cancer

Frozen section

A B S T R A C T

Background: This study investigated the frequency of high-risk cancer types in hysterectomy material obtained from patients who were diagnosed with atypical endometrial hyperplasia (AEH) by endometrial sampling. Materials and methods: A total of 227 patients with AEH were retrospectively included in the study. Hysterectomy material was examined as both perioperative frozen section (FS) and paraﬃn-embedded per-manent section (PS). Grade III tumors, grade II tumors larger than 2 cm, over 50% myometrial invasion, cervical involvement, and serous or clear cell histology were considered high-risk.

Results: Inﬁnal pathology, 57 (25.1%) patients had endometrial cancer and 7 (3%) patients had high-risk cancer. Overall analysis of FS/PS agreement yielded a Cohen's Kappa (K) coeﬃcient of 0.420 (moderate agreement). There was moderate (K = 0.526) agreement between FS and PS in detecting tumor grade, and good agreement (K = 0.653) in evaluation of myometrial invasion.

Conclusion: High-risk endometrial cancer can coexist with AEH. It should be remembered that despite pre-operative and FS examinations, these high-risk tumors can be overlooked untilﬁnal pathology.

1. Introduction

Endometrial hyperplasia refers to precancerous lesions character-ized by disordered proliferation of the endometrial glands due to chronic exposure to estrogen not counteracted by progesterone, and accounts for 15% of perimenopausal abnormal uterine bleeding [1,2]. In 1994, the World Health Organization (WHO) classified en-dometrial hyperplasia into four categories: simple hyperplasia without atypia, complex hyperplasia without atypia, simple hyperplasia with atypia, and complex hyperplasia with atypia [3]. This classification was revised in 2015 to include only two categories: hyperplasia without atypia, and atypical hyperplasia [4]. The rate of progression from atypical endometrial hyperplasia (AEH) to cancer is 8% for simple atypical hyperplasia and 29% for complex atypical hyperplasia [5]. The coexistence of endometrial cancer with AEH has been reported at rates of 20–50% in numerous studies [6–10]. Due to the potential for pro-gression to cancer and coexistence with cancer, hysterectomy is the definitive treatment of AEH [11]. Cancers coexisting with AEH are usually well-differentiated, superficially invasive endometrial carci-nomas that do not require lymphadenectomy [10]. However, surgical staging including lymphadenectomy is necessary for high-risk (serous

or clear cell histology, Grade III tumor, grade II tumors > 2 cm, > 50% myometrial invasion, cervical involvement) endometrial carcinomas [12].

This study was conducted to determine the frequency of high-risk lesions in hysterectomy material obtained from patients diagnosed with AEH by endometrial sampling.

2. Materials and methods

This single center retrospective study was conducted in accordance with the principles of the Declaration of Helsinki. We routinely request consent from patients we are going to operate on at our clinic, therefore written informed consent was obtained from each participant. Ethical approval was not obtained because of the retrospective nature of the study. The work has been reported in line with the STROCSS criteria [13].

A total of 248 patients with AEH who underwent hysterectomy in the Gynecological Oncology department of a research hospital between March 2004 and November 2017 were retrospectively evaluated. Twenty-one patients whose records were not available were excluded, and the remaining 227 patients were included in the study. AEH

https://doi.org/10.1016/j.ijsu.2018.04.019

Received 13 February 2018; Received in revised form 22 March 2018; Accepted 10 April 2018

∗_{Corresponding author. Department of Gynecologic Oncology, Kanuni Sultan Süleyman Training and Research Hospital, 34000, Istanbul, Turkey.}

E-mail address:drberdem@yandex.com(B. Erdem).

International Journal of Surgery 53 (2018) 350–353

Available online 14 April 2018

(2)

diagnosis was based on endometrial biopsy (EMB) performed using dilatation and curettage (D&C) or pipelle. Patients underwent surgery within 6 weeks of diagnosis. Hysterectomy material was examined as both perioperative frozen section (FS) and paraﬃn-embedded perma-nent section (PS). Based on FS results, surgical staging including lym-phadenectomy was performed to diagnose high-risk endometrial cancer. Grade III tumors, grade II tumors larger than 2 cm, over 50% myometrial invasion, cervical involvement, and serous or clear cell histology were considered high-risk. Tumors with none of these features were considered low-risk tumors. Non-endometroid tumors were eval-uated as grade III in both FS andﬁnal pathology.

Patients whose preoperative endometrial biopsy results indicated ‘AEH with coexistent endometrial cancer’ were excluded from the study, as were any endometrial hyperplasia or endometrial cancer pa-tients who had previously received medical treatment.

The patients' general characteristics and histopathological features including FS and PS results were recorded. Any FS and PS examination result that did not include hyperplasia or malignancy (polyp, dis-ordered proliferative endometrium, etc.) was accepted as benign. 2.1. Statistical analysis

All statistical analyses were performed using MedCalc software. P values < 0.05 were considered statistically signiﬁcant. Data are pre-sented as mean ± standard deviation (SD). Theχ2_{test and Student's}

t-test were used for comparative statistical analysis of unpaired data. Sensitivity, speciﬁcity, positive predictive values (PPVs), negative pre-dictive values (NPVs), and diagnostic accuracy were calculated with associated 95% conﬁdence intervals (CIs). The agreement between FS

diagnosis and diagnosis given in final pathological examination was calculated using Cohen's Kappa (κ) coefficient. Agreement rates be-tween FS and PS in the diagnosis of low- and high-risk endometrial cancer was also calculated. Theχ2test was used to compute odds ratios with 95% CIs for associations between each risk factor and endometrial cancer. Factors identified as potential risk factors in unadjusted ana-lyses (p < 0.05) were used to create a logistic regression model in which endometrial cancer was the dependent variable. Continuous variables were transformed into categorical variables using approx-imate optimal cut-off points identified by receiver operating char-acteristic (ROC) curve analysis.

3. Results

The general characteristics of the 227 patients included in the study are shown inTable 1. The patients' mean age was 49.4 (31–77) years, and the mean parity was 3.1 (0–11). Seventy percent of the patients had diabetes mellitus (DM). Preoperatively, 137 (60.3%) patients were di-agnosed by D&C with atypical hyperplasia, whereas endometrial cancer was detected in 57 (25.1%) patients inﬁnal pathology.

Comparison of the overall agreement between FS and PS examina-tions is shown inTable 2. FS examination revealed endometrial cancer in 35 patients (30 low-risk and 5 high-risk); in PS examination, 57 patients were diagnosed with endometrial cancer (50 low-risk, 7 high-risk). Overall analysis of FS/PS agreement yielded aκ coeﬃcient of 0.420 (moderate agreement).

The histopathological features of the endometrial cancers identiﬁed in PS examination are shown inTable 3. Endometroid cancer was de-tected in 54 patients (94.7%) and non-endometroid (clear cell + serous) cancer was observed in 3 patients (5.3%). The majority (70%) of patients had grade I cancers. Myometrial invasion greater than 50% was observed in only 3 patients (5.3%).

FS/PS agreement in tumor staging is shown inTable 4. In a total of 33 patients, both FS and PS were evaluated as cancer. There was moderate (K = 0.526) agreement between FS and PS in detecting tumor

Table 1

Clinical characteristics of the study population.

Charecteristic n:227 Age, years 49.4(31–77) Gravidy 4.1(0–15) Parity 3.1(0–11) Nulliparity 15(6.6) BMI, kg/m2 32.2(19–57) Hypertension 45(19.8) Diabetes mellitus 70(30.8) Smoking 39(17.2) Method of diagnosis D&C 137(60.3) Pipelle biopsy 90(39.7) Type of surgery

Only Total Hysterectomy ± BSO 212(93.3)

Staging surgery 15(6.6)

Number of removed lymph nodes 18.6(7–35)

Final pathology

Endometrial cancer 57(25.11)

No Malignancy 170(74.8)

Values are expressed as means (with ranges) or numbers (with percen-tages); number of removed lymph nodes is expressed as median (with range).

Table 2

Results of FS versus PS analyses.

Paraﬁn

Benign(51) Hyperplasia without atypi(58) Hyperplasia with atypi(61) Low risk EC(50) High risk EC(7)

Frozen Benign(123) 49 42 26 5 1

Hyperplasia without atypi(23) 2 16 4 1 0

Hyperplasia with atypi(46) 0 0 29 16 1

Low risk EC(30) 0 0 2 26 2

High risk EC(5) 0 0 0 2 3

FS; frozen section, PS:paraﬁn section. κ coeﬃcient: 0.420.

Table 3

Histopathological characteristics of EC as revealed in PS examination. Charecteristic n:57 Histopathological feature Endometrioid 54(94.7) Serous 1(1.7) Clear cell 2(3.5) Grade I 40(70) II 13(22.8) III 4(7) Myometrialİnvasion < 50 54(94.7) ≥50 3(5.3)

Values are expressed as numbers (with percentages) or mean (with range).

B. Erdem et al. International Journal of Surgery 53 (2018) 350–353

(3)

grade, and good agreement (K = 0.653) in evaluation of myometrial invasion (seeTable 5).

Older age (p < 0.001), higher body mass index (p = 0.001), DM (p = 0.02), hypertension (HT) (p = 0.017), and nulliparity (p < 0.001) were identiﬁed as risk factors for coexistence of en-dometrial cancer with AEH (Table 6).

Logistic regression analysis revealed that age over 50 years, DM, HT, and nulliparity were independent risk factors endometrial cancer coexisting with AEH (Table 7).

4. Discussion

It is not always easy to separate AEH, which is a precursor for en-dometrial cancer, from well-diﬀerentiated endometrial carcinoma [11]. Zaino et al. reported that agreement between pathologists is low (38–47%) in the diagnosis of AEH [14]. The likelihood of encountering endometrial cancer in AEH diagnosed with D&C specimens is lower than in those diagnosed with pipelle biopsy (33% versus 47%) [15].

In their literature review including 2572 patients, Rakha et al. de-termined that endometrial carcinoma was detected in later biopsies or hysterectomy material in 37% of patients diagnosed with AEH from an endometrial sample [16]. Giede et al. reported that the prevalence of concurrent carcinomas in AEH patients was 35.7% [17], while Miller et al. [18] and Whyte et al. [19] reported that this rate as 46% and 28.4% (25/88), respectively.

Hahn et al. evaluated the deﬁnitive post-hysterectomy pathology of

126 patients with AEH, and endometrial cancer was detected in 13 patients (10.3%). This low rate was attributed to the lower incidence of endometrial cancer in the Asian population and higher preoperative detection rates [20]. In our study, this rate was 25.1%, which is in accordance with the literature.

The presence of high-risk tumors necessitates staging surgery in-cluding lymph node dissection. High-risk tumors can be seen in 5–7% of all AEH patients [21,22].

Leitao et al. evaluated 197 patients with atypical complex hyper-plasia and found that 34% had underlying endometrial cancer. Tumor type was endometroid in 90%, not otherwise speciﬁed carcinoma in 7.5%, serous in 1.5%, and metastatic breast carcinoma in 1.5% of those patients. Ninety-one percent of the cancers were grade I, while 6% were grade II and 3% were not assessed (serous + metastatic breast carci-noma) [23]. Hahn et al. reported that all of the cancers coexisting with AEH were endometroid adenocarcinomas and none showed myometrial or lymphovascular invasion. Eighty-four percent of the cancers were grade I and 15% were grade II [20]. Whyte et al. detected high-risk cancer in 5.6% (n = 5) patients in a study including 88 AEH patients [19]. In our study,ﬁnal pathology results indicated high-risk tumor in 3% (7/227) of AEH patients, who then underwent staging surgery with pelvic/paraaortic lymphadenectomy.

FS examination is controversial in patients with AEH, and FS/PS agreement is 60–90% [21,22]. In patients with endometrial cancer, agreement between FS and PS was 98.6% for histologic subtype, 84.3% for grade, and 94.3% for myometrial invasion [22]. Leitao et al. com-pared the FS and PS results of 482 grade I endometrial adenocarci-nomas. Grade inﬁnal pathology was higher than the grade determined by intraoperative FS in 15% of the patients [24]. In contrast, Kumar et al. reported only 1.3% discordance between FS andﬁnal pathology [25]. We observed moderate agreement between FS and PS in detection of malignancy in the present study. In two patients, FS was evaluated as endometrial cancer and PS was evaluated as atypical hyperplasia. The main reason for the poor agreement in our study was the detection of endometrial cancer in PS of 17 of the 46 patients considered atypical hyperplasia on FS. There was moderate concordance between FS and PS in determining tumor grade and good concordance for myometrial in-vasion.

Limitations of this study are its retrospective design and that 39.7% of the endometrial samples were obtained using a pipelle. If all samples

Table 4

Tumour grade on frozen section reporting compared toﬁnal histopathology of the uterus for EC.

Paraﬁn Section Frozen Section

Grade I Grade II Grade III

Grade I 19 0 0

Grade II 6 3 1

Grade III 0 1 3

Total 25(75%) 4(12%) 4(12%)

Values are given as n (%). κ coeﬃcient 0.526.

Table 5

The depth of tumour invasion on frozen section reporting compared to theﬁnal histopathology of the uterus.

Paraﬁn Frozen < 50 MI ≥50 MI < 50 MI 31 0 ≥50 MI 1 1 Total 32(97.0) 1(3.0) MI:myometrial invasion. κ coeﬃcient: 0.653. Table 6

Unadjusted odds ratios, using demographic factors and preoperative characteristics as risk factors, for EC presence in patients with AEH.

Factor AEH without EC(n:170) AEH with EC(n:57) p-value OR

Age ± SD(year) 48.00 ± 6.5 53.5 ± 8.8 < 0,001

Tobacco use, n (%) 30(17.6) 9(15.8) 0.748 1.1(0.5–2.0)

Nulliparity, n(%) 8(4.6) 7(12.7) < 0.001 0.3(0.1–0.9)

Body mass index ± SD*(kg/m2) 31.5 ± 5.9 34.6 ± 6.9 0.001

DM,n(%) 28(16.5) 17(29.8) 0.029 0.5(0.3–0.9)

HT,n(%) 45(26.5) 25(43.9) 0.017 0.5(0.3–0.8)

Values are expressed as numbers (with percentages). Odds ratios are presented with 95% CI in parentheses.

AEH: atypical endometrial hyperplasia, EC: Endometrial cancer, DM:Diabetes mellitus, HT:hypertension, SD:Standard deviation. Table 7

Odds ratios derived using the logistic regression model with EC as the depen-dent variable.

Factor Odds Ratio pValue

Age > 50 yearsa _{3.9(1.8–8.3)} _0.0003

BMI > 30 kg/m2a _{1.0(0.5–2.2)} _0.8082

DM 2.0(1.0–4.1) 0.0041

HT 2.6(1.3–5.1) 0.0022

Nulliparity 3.5(1.6–8.1) 0.0029

Values in parentheses are 95% CI.

a _{According to ROC analysis.}

(4)

were taken with D & C, our endometrial cancer rate could be lower. In summary, although most concurrent endometrial cancers in pa-tients with AEH are well-differentiated endometroid cancer, high-risk endometrial cancers may also be encountered, as demonstrated in this study. It should be taken into consideration that despite preoperative and FS examinations, these high-risk tumors may be overlooked until final pathology. Larger prospective trials are needed to achieve a defi-nite answer as to can concurrent high risc endometrial carcinoma occur with the atypical endometrial carcinoma?

Ethical approval

Because of the retrospective nature of the study, we do not have ethical approval but institutional approval was obtained. The retro-spective study was conducted in accordance with the principles of the Declaration of Helsinki. A written informed consent was obtained from each participant.

Sources of funding

This research did not receive any speciﬁc grant from funding agencies in the public, commercial, or not-for-proﬁt sectors.

Author contribution

Baki Erdem; data collections, study design, writing, surgery. Nuri Peker; study design, writing.

Osman aşıcıoğlu; statistical analysis. Niyazi alper seyhan; data collection. Volkan Ülker; surgery, study design, writing. Özgür Akbayır; surgery, study design, writing. Conﬂicts of interest

The authors have no conﬂict of interests.

Research registration Unique Identifying Number (UIN) Researchregistry 3686.

Guarantor Baki erdem. References

[1] L. Hedrick Ellenson, B.M. Ronnett, R.J. Kurman, Precursor lesions of endometrial carcinoma, in: R.J. Kurman, L. Hedrick Ellenson, B.M. Ronnett (Eds.), Blaustein's Pathology of the Female Genital Tract, sixth ed., Springer, New York, 2010, pp. 360–361.

[2] S. Feldman, E.F. Cook, B.L. Harlow, R.S. Berkowitz, Predicting endometrial cancer among older women who present with abnormal vaginal bleeding, Gynecol. Oncol. 56 (3) (1995) 376–381.

[3] R.E. Scully, T.A. Bonﬁglio, Kurman, et al., Uterine corpus, Histological Typing of

Female Genital Tract Tumours, second ed., Springer-Verlag, New York, 1994, p. 13. [4] G. Emons, M.W. Beckmann, D. Schmidt, et al., New WHO classiﬁcation of

en-dometrial hyperplasias, Geburtshilfe Frauenheilkd 75 (2015) 135.

[5] R.J. Kurman, P.F. Kaminski, H.J. Norris, The behavior of endometrial hyperplasia: a long-term study of“untreated” hyperplasia in 170 patients, Cancer 56 (1985) 403–412.

[6] J.E. Hunter, D.E. Tritz, M.G. Howell, P.D. DePriest, H.H. Gallion, S.J. Andrews, et al., The prognostic and therapeutic implications of cytologic atypiain patients with endometrial hyperplasia, Gynecol. Oncol. 55 (1994) 66–71.

[7] E.A. Widra, C.J. Dunton, M. McHugh, J.P. Palazzo, Endometrial hyperplasia and the risk of carcinoma, Int. J. Gynecol. Canc. 5 (1995) 233–235.

[8] M.F. Janicek, N.B. Rosenshein, Invasive endometrial cancer in uteri resected for atypical endometrial hyperplasia, Gynecol. Oncol. 52 (1994) 373–378. [9] T. Bilgin, S. Ozuysal, H. Ozan, T. Atakan, Coexisting endometrial cancer in patients

with a preoperative diagnosis of atypical endometrial hyperplasia, J. Obstet. Gynaecol. Res. 30 (2004) 205–209.

[10] C.L. Trimble, J. Kauderer, R. Zaino, S. Silverberg, P.C. Lim, J.J. Burke 2ndet al., Concurrent endometrial carcinoma in women with a biopsy diagnosis of atypical endometrial hyperplasia: a Gynecologic Oncology Group study, Cancer 106 (2006) 812–819.

[11] R.J. Kurman, H.J. Norris, Evaluation of criteria for distinguishing atypical en-dometrial hyperplasia from well-diﬀerentiated carcinoma, Cancer 49 (1982) 2547–2559.

[12] A. Mariani, S.C. Dowdy, W.A. Cliby, et al., Prospective assessment of lymphatic disseminationin endometrial cancer: a paradigm shift in surgicalstaging, Gynecol. Oncol. 109 (2008) 11–18.

[13] R.A. Agha, M.R. Borelli, M. Vella-Baldacchino, R. Thavayogan, D.P. Orgillfor the STROCSS Group, The STROCSS statement: strengthening the reporting of cohort studies in surgery, Int. J. Surg. 46 (2017) 198–202.

[14] R.J. Zaino, J. Kauderer, C.L. Trimble, S.G. Silverberg, J.P. Curtin, P.C. Lim, D.G. Gallup, Reproducibility of the diagnosis of atypical hyperplasia.A gynecologic Oncology group study, Cancer 106 (4) (2006) 804–811.

[15] E. Suh-Burgmann, Y. Hung, M.A. Armstrong, Complex atypical hyperplasia: the risk of unrecognized adenocarcinoma and value for preoperative dilation and curettage, Obstet. Gynecol. 114 (2009) 523–529.

[16] E. Rakha, S.C. Wong, I. Soomro, et al., Clinical outcome of atypical endometrial hyperplasia diagnosed on an endometrial biopsy: institutional experience and re-view of literature, Am. J. Surg. Pathol. 36 (2012) 1683.

[17] K.C. Giede, T.W. Yen, R. Chibbar, R.A. Pierson, Signiﬁcance of concurrent en-dometrial cancer in women with a preoperative diagnosis of atypical enen-dometrial hyperplasia, J. Obstet. Gynaecol. Can. 30 (2008) 896–901.

[18] C. Miller, M.A. Bidus, J.P. Pulcini, G.L. Maxwell, J.A. Cosin, G.S. Rose, The ability of endometrial biopsies with atypical complex hyperplasia to guide surgical man-agement, Am. J. Obstet. Gynecol. 199 (1) (2008) 69e1–69e4.

[19] J.S. Whyte, E.P. Gurney, J.P. Curtin, S.V. Blank, Lymph node dissection in the surgical management of atypical endometrial hyperplasia, Am. J. Obstet. Gynecol. 202 (2) (2010) 176e1–176e4.

[20] H.S. Hahn, Y.K. Chun, Y.I. Kwon, T.J. Kim, K.H. Lee, J.U. Shim, J.E. Mok, K.T. Lim, Concurrent endometrial carcinoma following hysterectomy for atypical en-dometrial hyperplaisa, Eur. J. Obstet. Gynecol. Reprod. Biol. 150 (1) (2010) 80–83. [21] M.D. Indermaur, B. Shoup, S. Tebes, J.M. Lancester, The accuracy of frozen

pa-thology at time of hysterectomy in patients with complex atypical hyperplasia on preoperative biopsy, Am. J. Obstet. Gynecol. 196 (5) (2007) e40–e42. [22] S. Attard Montalto, M. Coults, O. Devaja, J. Summers, R. Jyothirmayi,

A. Papadopoulos, Accuracy of frozen section diagnosis at surgery in pre- malignant and malignant lesions of the endometrium, Eur. J. Gynaecol. Oncol. 29 (5) (2008) 435–440.

[23] M.M. Leitao Jr., G. Han, L.X. Lee, N.R. Abu-Rustum, C.L. Brown, D.S. Chi, Y. Sonoda, D.A. Levine, G.J. Gardner, E.E. Jewell, R.R. Barakat, R.A. Soslow, Complex atypical hyperplasia of the uterus: characteristics and prediction of un-derlying carcinoma risk, Am. J. Obstet. Gynecol. 203 (4) (2010) 349 e1-6. [24] M.M. Leitao Jr., S. Kehoe, R.R. Barakat, et al., Accuracy of preoperative endometrial

sampling diagnosis of FIGO grade 1 endometrial adenocarcinoma, Gynecol. Oncol. 111 (2008) 244.

[25] S. Kumar, F. Medeiros, S.C. Dowdy, et al., A prospective assessment of the reliability of frozen section to direct intraoperative decision making in endometrial cancer, Gynecol. Oncol. 127 (2012) 525.