Sei sulla pagina 1di 7

Original Research ajog.

org

GYNECOLOGY
Preoperative predictors of endometrial cancer at time of
hysterectomy for endometrial intraepithelial neoplasia
or complex atypical hyperplasia
Monica Hagan Vetter, MD; Blair Smith, MD; Jason Benedict, MS; Erinn M. Hade, PhD; Kristin Bixel, MD; Larry J. Copeland, MD;
David E. Cohn, MD; Jeffrey M. Fowler, MD; David O’Malley, MD; Ritu Salani, MD, MBA; Floor J. Backes, MD

BACKGROUND: Endometrial intraepithelial neoplasia, also known as the time of hysterectomy was developed using logistic regression with 5-
complex atypical hyperplasia, is a precancerous lesion of the endometrium fold cross-validation.
associated with a 40% risk of concurrent endometrial cancer at the time of RESULTS: Of the 1055 charts reviewed, 169 patients were eligible and
hysterectomy. Although a majority of endometrial cancers diagnosed at included. Of these patients, 87 (51.5%) had a final diagnosis of endo-
the time of hysterectomy for endometrial intraepithelial neoplasia are low metrial intraepithelial neoplasia/other benign disease, whereas 82
risk and low stage, approximately 10% of patients ultimately diagnosed (48.5%) were ultimately diagnosed with endometrial cancer. No medical
with endometrial cancers will have high-risk disease that would warrant comorbidities were found to be strongly associated with concurrent
lymph node assessment to guide adjuvant therapy decisions. Given these endometrial cancer. Patients with endometrial cancer had a thicker
risks, some physicians choose to refer patients to a gynecologic oncologist average endometrial stripe compared to the patients with no endometrial
for definitive management. Currently, few data exist regarding preoper- cancer at the time of hysterectomy (15.7 mm; standard deviation, 9.5)
ative factors that can predict the presence of concurrent endometrial versus 12.5 mm; standard deviation, 6.4; P ¼ .01). An endometrial stripe
cancer in patients with endometrial intraepithelial neoplasia. Identification of 2 cm was associated with 4.0 times the odds of concurrent endo-
of these factors may assist in the preoperative triaging of patients to metrial cancer (95% confidence interval, 1.5e10.0), controlling for age. In
general gynecology or gynecologic oncology. all, 87% of endometrial cancer cases were stage T1a (Nx or N0).
OBJECTIVE: To determine whether preoperative factors can predict the Approximately 44% of patients diagnosed with endometrial cancer and an
presence of concurrent endometrial cancer at the time of hysterectomy in endometrial stripe of 2 cm met the “Mayo criteria” for indicated lym-
patients with endometrial intraepithelial neoplasia; and to describe the phadenectomy compared to 22% of endometrial cancer patients with an
ability of preoperative characteristics to predict which patients may be at a endometrial stripe of <2 cm.
higher risk for lymph node involvement requiring lymph node assessment CONCLUSION: Endometrial stripe thickness and age were the stron-
at the time of hysterectomy. gest predictors of concurrent endometrial cancer at time of hysterectomy
MATERIALS AND METHODS: We conducted a retrospective for endometrial intraepithelial neoplasia. Referral to a gynecologic
cohort study of women undergoing hysterectomy for pathologically oncologist may be especially warranted in endometrial intraepithelial
confirmed endometrial intraepithelial neoplasia from January 2004 to neoplasia patients with an endometrial stripe of 2 cm given the
December 2015. Patient demographics, imaging, pathology, and out- increased rate of concurrent cancer and potential need for lymph node
comes were recorded. The “Mayo criteria” were used to determine assessment.
patients requiring lymphadenectomy. Unadjusted associations between
covariates and progression to endometrial cancer were estimated by 2- Key Words: endometrial cancer, endometrial hyperplasia, endometrial
sample t-tests for continuous covariates and by logistic regression for intraepithelial neoplasia, endometrial stripe, precancerous lesion, trans-
categorical covariates. A multivariable model for endometrial cancer at vaginal pelvic ultrasound

E ndometrial intraepithelial neoplasia


(EIN), formally known as complex
atypical hyperplasia (CAH), is a prema-
of clinical significance because of an
approximately 40% risk of progression to
endometrial cancer (EC).1e3 Further-
hysterectomy, with or without bilateral
salpingo-oophorectomy. This treatment
protocol allows for full pathologic evalu-
lignant lesion of the endometrium that is more, the prevalence of concurrent EC in ation and assessment of concurrent can-
patients diagnosed with EIN undergoing cer, and provides definitive therapy.4,9
hysterectomy approaches 43%.4 Risk Nonsurgical management may be
Cite this article as: Vetter MH, Smith B, Benedict J, et al.
Preoperative predictors of endometrial cancer at time of factors for the development of EIN appropriate for patients desiring future
hysterectomy for endometrial intraepithelial neoplasia or include obesity, anovulation, nulliparity, fertility or for those patients with
complex atypical hyperplasia. Am J Obstet Gynecol and diabetes.5,6 A diagnosis of EIN can be comorbidities precluding surgical
2020;222:60.e1-7. made by outpatient endometrial biopsy management.10
0002-9378/$36.00 or by dilation and curettage, with or The majority of EIN patients
ª 2019 Elsevier Inc. All rights reserved. without hysteroscopy.7,8 Given the high ultimately diagnosed with EC will
https://doi.org/10.1016/j.ajog.2019.08.002
risk of concurrent cancer and the risk of have early stage, low-risk disease.
progression, the standard treatment of However, approximately 12% will have
EIN is surgical management with high-grade tumors with deep

60.e1 American Journal of Obstetrics & Gynecology JANUARY 2020


ajog.org GYNECOLOGY Original Research

second surgery based on intra-


AJOG at a Glance operative frozen section, final pa-
Why was this study conducted? thology diagnosis, and/or provider
To determine whether there are preoperative predictors of concurrent uterine choice. Patients with no preoperative
adenocarcinoma at time of hysterectomy for endometrial intraepithelial imaging or those who had undergone
neoplasia, and need for lymph node assessment at time of surgical management. prehysterectomy endometrial ablation
were excluded. Patient demographics,
Key findings imaging results, pathologic data, and
Both preoperative endometrial stripe thickness and smoking status predicted the outcomes were recorded from the
presence of concurrent cancer at time of hysterectomy for endometrial intra- medical record.
epithelial neoplasia. Endometrial stripe thickness of 2 cm was associated with a We used the widely accepted “Mayo
4-fold increase in odds of concurrent cancer compared to endometrial stripe criteria” to calculate for whom a
thickness of <2 cm. In all, 44% of patients with an endometrial stripe of 2 cm lymph node dissection would be rec-
met clinical criteria for lymph node assessment. ommended.15 The Mayo criteria
recommend a lymph node dissection
What does this add to what is known? for patients with the following: grade
This study suggests that referral to gynecologic oncology may be warranted for 1 or grade 2 endometrioid adenocar-
patients with endometrial intraepithelial neoplasia, especially if endometrial cinoma 2 cm and >50% myometrial
stripe thickness is 2 cm, given the increased rate of concurrent cancer and the invasion, any grade 3 endometrioid
potential need for lymph node assessment. adenocarcinomas, and all non-
endometrioid adenocarcinomas (se-
rous, clear cell, mixed, carcinosar-
myometrial invasion and a 3e7% risk suspicious feature as important in the coma). Descriptive statistics (counts,
of lymph node involvement.4,11e13 prediction of underlying cancer risk, frequency, mean, and standard devia-
Although the comprehensive surgical with the highest risk in patients assigned tion) were reported. Initial associa-
staging with lymph node assessment via the designation of EIN suspicious.16 tions with progression to EC were
full lymphadenectomy or sentinel However, few data exist on the impact made by 2-sample t tests for contin-
lymph node approach for all patients of objective preoperative factors that uous covariates and by the estimation
with EIN would result in overtreatment may be used to predict the risk of un- of odds ratios (OR), and 95% confi-
in a large proportion of patients, there derlying EC. dence intervals (CI), for categorical
remains a subset of patients for whom The primary purpose of this study covariates. A predictive multivariable
lymph node assessment as a guide to was to determine whether preoperative logistic model for progression to EC
adjuvant therapy is beneficial in factors, including imaging and patient was developed by forward selection of
reducing the risk of over- or under- characteristics, can predict the pres- covariates with crude association at
treatment. In addition, hysterectomy ence of concurrent EC at the time of the 10% level and 5-fold cross-
results in disruption of the lymphatic hysterectomy in patients diagnosed validation. Model discrimination was
channels, making sentinel lymph node with EIN. A secondary outcome was assessed by the area under the receiver
assessment impossible to perform after to describe the ability of preoperative operating characteristic curve (AUC),
hysterectomy, in the event of an EC characteristics to predict which pa- and model calibration was assessed
diagnosis on intraoperative or final pa- tients may require a lymph node using the HosmerLemeshow
thology. This fact has resulted in dissection. goodness-of-fit test. The associations
ongoing discussions about whether or between the Mayo criteria and either
not a referral to a gynecologic oncolo- Materials and Methods LND, lymphadenectomy, or EMS
gist is warranted in all cases of EIN.14 After obtaining institutional review thickness were assessed using the
Given the challenging management board approval, a retrospective chart Fisher exact test. All reported P values
decisions associated with EIN, interest review was performed, encompassing are 2 sided. Data were analyzed in
exists in identifying factors that may all patients undergoing hysterectomy Stata version 15.1 (StataCorp, College
improve preoperative risk prediction of and bilateral salpingo-oophorectomy Station, TX).
EC. Previous studies have identified for confirmed EIN at The Ohio
sampling method as being associated State University from January 2004 to Results
with EC risk, with EIN diagnosed on December 2015. Confirmation of an In total, 169 individuals were eligible to
office biopsy alone being more strongly EIN diagnosis by an Ohio State Uni- be included in the study. EIN was diag-
associated with EC on follow-up.15,16 A versity pathologist was required for nosed by endometrial biopsy in 44% of
retrospective study identified pathologic inclusion. Lymph node assessment the patients, whereas EIN was initially
characteristics such as extent of EIN or was performed either at the time of diagnosed by dilation and curettage in
involvement of a polyp or other primary surgery or, in rare cases, in a 56% of the patients. In all, 73 patients

JANUARY 2020 American Journal of Obstetrics & Gynecology 60.e2


Original Research GYNECOLOGY ajog.org

TABLE 1
Baseline patient characteristics by final pathology diagnosis at time of hysterectomy
Endometrial cancer Benign Pathology Unadjusted OR Adjusted ORa
Characteristic (n ¼ 82), n (%) (n ¼ 87), n (%) (95% CI) (95% CI)
EMS thickness
<2 cm 59 (42.4) 80 (57.6) Reference Reference
2 cm 23 (76.7) 7 (23.3) 4.5 (1.8e11.1) 4.0 (1.6e10.1)
b
Smoking status
Never smoker 55 (52.9) 49 (47.1) Reference
Ever smoker 25 (41.7) 35 (58.3) 0.6 (0.3e1.2)
Age, y
50 23 (42.6) 31 (57.4) Reference Reference
51e64 39 (45.9) 46 (54.1) 1.1 (0.6e2.3) 1.2 (0.6e2.4)
65 20 (66.7) 10 (33.3) 2.7 (1.1e6.8) 2.3 (0.9e5.9)
Race/ethnicity
Nonwhite 7 (53.8) 6 (46.2) Reference
White 75 (48.1) 81 (51.9) 0.8 (0.3e2.5)
c
Menopausal status
Premenopausal 25 (41.0) 36 (59.0) Reference
Postmenopausal 56 (52.8) 50 (47.2) 1.6 (0.9e3.0)
Comorbidities
BMI
Not obese (<30) 14 (48.3) 15 (51.7) Reference
Obesed (30) 68 (48.6) 72 (51.4) 1.0 (0.5e2.3)
Hypertension
No 25 (44.6) 31 (55.4) Reference
Yes 57 (50.4) 56 (49.6) 1.3 (0.7e2.4)
Diabetes
No 58 (47.5) 64 (52.5) Reference
Yes 24 (51.1) 23 (48.9) 1.2 (0.6e2.3)
BMI, body mass index (kg/m2); CI, confidence interval; EC, endometrial cancer; EMS, endometrial stripe thickness; OR, odds ratio.
a
Adjusted odds ratio includes the variables in the final model only; b Smoking status has 5 missing values (2 in EC group and 3 in benign pathology group); c Menopausal status has 2 missing values
(1 in EC group and 1 in benign pathology group); d Obesity was defined as BMI 30.
Vetter et al. Predictors of cancer at time of surgical management of endometrial intraepithelial neoplasia. Am J Obstet Gynecol 2020.

(43%) had a final diagnosis of EIN; 82 those with EIN/other benign disease. maximum: 21.0, 64.3) for those pro-
(48%) were diagnosed with endometrial As women aged, they were more often gressing to EC and 41.5 (minimum,
cancer at the time of hysterectomy, diagnosed with EC (65 years vs 50 maximum: 19.9, 69.2) for those who
whereas 14 (8%) had other benign years: OR, 2.7; 95% CI, 1.1e6.8), P ¼ did not. The rate of hypertension,
disease. .08). In both groups, most patients diabetes, and breast cancer were similar
Demographics for the study popu- were of white race/ethnicity, were between the 2 groups.
lation are reported in Table 1 according postmenopausal, and had never used The final multivariable model
to final pathologic diagnosis. The hormone replacement therapy. The (Figure 1) found preoperative trans-
average age of patients diagnosed with most commonly noted medical co- vaginal ultrasound endometrial stripe
EC at the time of hysterectomy was 56 morbidity in both groups was obesity (EMS) and age group to be strongly
years (standard deviation [SD], 10.0), (83% in both groups). Median body associated with an increased odds of
compared to 54 years (SD, 10.0) for mass index was 39.5 (minimum, EC. The final predictive model had

60.e3 American Journal of Obstetrics & Gynecology JANUARY 2020


ajog.org GYNECOLOGY Original Research

histology, grade 3, and was noted to have


FIGURE 1
an EMS of 13.5 mm prior to
Multivariable logistic regression model for endometrial cancer at time of
hysterectomy.
hysterectomy
In the entire cohort, there was a
single recurrence after a median
follow-up of 2.4 years. This patient
had stage IA, grade I endometrioid EC
with negative lymphovascular space
invasion. She did not undergo lym-
phadenectomy and did not receive
adjuvant therapy. She was found to
have pelvic and nodal recurrences 12
months after hysterectomy. She
declined any additional therapy in
favor of hospice placement, and died
of disease within 2 months.

Comment
Principal findings
Our study demonstrates that patients
Vetter et al. Predictors of cancer at time of surgical management of endometrial intraepithelial neoplasia. Am J Obstet
Gynecol 2020. with a preoperative diagnosis of complex
endometrial hyperplasia and an EMS of
2 cm have 4.0 (95% CI: 1.6, 10.1) times
reasonable discrimination (AUC, 0.64; Approximately 24% (n ¼ 20) of EC the odds of endometrial cancer at the
95% CI, 0.56e0.72) and calibration patients underwent lymphadenectomy; time of hysterectomy, controlling for
(HosmerLemeshow goodness-of-fit 7 of these cases (35%) met Mayo age. Furthermore, 44% (n¼ 10) of pa-
test, P ¼ 0.41). Patients with EC at criteria. Of 82 patients with endome- tients with an EMS of 2 cm would have
the time of hysterectomy had an trial cancer, 23 (28%) met Mayo required lymph node dissection based
average EMS of 15.7 (SD, 9.5) mm criteria. Ten patients (44%) with an on Mayo criteria.
compared to those with EIN/other EMS of 2 cm met Mayo criteria on
benign disease 12.5 (SD, 6.4). Patients final pathology, compared to 13 pa- Clinical implications
with an EMS of 2 cm had 4.0 times tients (22%) with an EMS of <2 cm, Management of EIN continues to be a
the odds of concurrent EC (OR, 4.0; indicating that a lymph node dissec- complex issue, especially when deciding
95% CI, 1.6e10.1), controlling for age tion would be warranted. Of the pa- on referral to a gynecologic oncologist.
(Table 1). An increased odds of EC is tients with an EMS of 2 cm and EC Previous studies exploring the predic-
also suggested for those patients with diagnosis, 20 patients (87.0%) had tion of the risk of concurrent EC in pa-
increased age (65 years and older; grade 1, 2 patients (8.7%) had grade 2, tients with EIN have focused on factors
adjusted OR, 2.3, 95% CI, 0.9e5.9), and 1 patient (4.4%) had grade 3. Two such as sampling method and histologic
compared to patients 50 years and patients (8.7%) had nonendometrioid characteristics such as type and extent of
younger, controlling for EMS histology, 4 (17.4%) had >50% myo- EIN.16 For the first time, we have
(Table 1). metrial invasion, and 3 (13.0%) had demonstrated that the preoperative EMS
Table 2 describes the characteristics lymphovascular space invasion. determined by preoperative transvaginal
of patients who were diagnosed with There were 2 instances of stage III ultrasound is associated with increased
EC at the time of hysterectomy. The disease. One patient had a dediffer- odds of EC, while controlling for age.
majority of EC cases were early stage entiated carcinoma at the time of hys- These results may be particularly rele-
as defined as T1a (Nx or N0) (87%), terectomy and subsequently underwent vant to women and their gynecologists
and had low-risk pathologic features imaging, as she did not undergo lymph who are in settings that may be a distance
such as grade 1 disease (92%), endo- node assessment at the time of hyster- away from a gynecologic oncologist.
metrioid histology (96%), and lack of ectomy. Her imaging demonstrated Applying this assessment may assist in
lymphovascular space invasion (89%). lymphadenopathy that was biopsied and making the decision/plan for referral.
However, 10% of patients had >50% found to be consistent with metastatic Within the cohort diagnosed with
myometrial invasion; these patients disease. She then received systematic concurrent EC, the majority of cases
were considered to be stage IB or chemotherapy without complete lym- were early stage with low-risk patho-
higher. Of the EC cohort, 15% phadenectomy and had no evidence of logic features such as grade 1 or
received adjuvant therapy, most disease at 17 months. The other patient endometrioid histology. In fact, almost
commonly with vaginal brachytherapy. had stage IIIC disease of endometrioid half of these cases were confined to the

JANUARY 2020 American Journal of Obstetrics & Gynecology 60.e4


Original Research GYNECOLOGY ajog.org

endometrium. There was a lower rate


TABLE 2
of high-risk disease in our cohort
Clinicopathological characteristics of patients diagnosed with endometrial
(grade 2/3 disease or outer 50%
carcinoma at time of final pathology
myoinvasion) compared to those in
previously studied cohorts.4,17 This Characteristic Endometrial carcinoma (n ¼ 82), n (%)
may reflect demographic differences, Stage
such as a higher proportion of patients
IA 71 (86.6)
of white race/ethnicity and a higher
body mass index in our cohort, both IB 4 (4.9)
of which are associated with the II 5 (6.1)
development of type I endometrial III 2 (2.4)
cancers.18 Type I endometrial cancers
Grade
are most often estrogen dependent
and are associated with lower grades 1 75 (91.5)
and rates of myoinvasion, lympho- 2 4 (4.9)
vascular space invasion, and lymph 3 3 (3.7)
node involvement compared to type II
endometrial cancers.19 Histology
The major advantage in referring Endometrioid 79 (96.3)
patients with EIN to a gynecologic Mixed pattern 2 (2.4)
oncologist is the gynecologic oncolo- Dedifferentiated 1 (1.2)
gist’s ability to perform comprehensive
staging, including lymph node assess- LVSI
ment when needed. Although the Present 9 (11.0)
impact of routine lymphadenectomy on Absent 73 (89.0)
survival is controversial, the benefit of
Meets Mayo criteria 23 (28.0)
lymph node assessment lies within
determining which patients need adju- Underwent LND 20 (24.4)
vant therapy.20 Previously, full lym- Recurrence 1 (1.2)
phadenectomy was performed as part of LND, lymphadenectomy; LVSI, lymphovascular space invasion.
comprehensive surgical staging, but was Vetter et al. Predictors of cancer at time of surgical management of endometrial intraepithelial neoplasia. Am J Obstet
Gynecol 2020.
associated with increased lymphe-
dema.21 Interest then turned to defining
low-risk patients in whom full lymph
node dissection could be avoided. EC involves injecting the cervix with is to offer SLN dissection to all patients
Mariani et al defined a low-risk popu- indocyanine green a fluorescent dye, with EIN. This would allow for a less
lation consisting of patients with grade with or without a colorimetric dye or morbid, yet reliable, lymph node
1 or 2 endometrioid adenocarcinoma radioactive tracer, and has been shown assessment and avoidance of a second
with less than 50% myometrial invasion to be a reliable and safe alternative to surgical procedure and its associated
and a tumor <2 cm in diameter.19e22 A full lymphadenectomy, with a low risks in the event of an EC diagnosis.
prospective study using these criteria false-negative rate.23e27 It is important Based on our study, the risk of meeting
demonstrated a <1% risk of lymph to note that the ability to perform SLN Mayo criteria (indicating the need for
node involvement in these low-risk mapping in EC depends on intact, lymphadenectomy) in patients with
patients compared to 16% in those not unobstructed lymphatic channels EIN/CAH was 28%. SLN mapping
meeting the criteria. Notably, in our originating from the uterine corpus should be considered for all patients with
cohort, 44% of patients with an EMS of and cervix and cannot be performed EIN/CAH, especially for patients with an
2 cm would have required lymph after hysterectomy. EIN patients diag- endometrial stripe thickness of >2 cm
node dissection using the criteria nosed with EC with high-risk features on preoperative pelvic ultrasound.
described above. at the time of hysterectomy alone Knowledge of lymph node status in pa-
Most recently, sentinel lymph node would then subsequently require a full tients with high intermediate uterine risk
(SLN) mapping has been introduced lymphadenectomy. This results in the factors would allow a more tailored
into the surgical management of patient being exposed to additional recommendation for postoperative
endometrial cancer, with the intention anesthesia risks and surgical risks therapy or surveillance, with possible
to reduce morbidity associated with a associated with lymphadenectomy. omission of external beam radiation
full lymphadenectomy (eg, lymphe- One alternative approach to manage- therapy and/or chemotherapy for pa-
dema, lymphocele). SLN mapping in ment of EIN in the era of SLN mapping tients at low risk for recurrence.28,29

60.e5 American Journal of Obstetrics & Gynecology JANUARY 2020


ajog.org GYNECOLOGY Original Research

Strengths and limitations 2. Mutter GL. Endometrial intraepithelial hyperplasia: the risk of unrecognized adeno-
Limitations of this study include its neoplasia (EIN): will it bring order to chaos? The carcinoma and value of preoperative dilation
Endometrial Collaborative Group. Gynecol and curettage. Obstet Gynecol 2009;114:
retrospective nature and potential lack of Oncol 2000;76:287–90. 523–9.
measured confounders. In addition, the 3. Baak JP, Mutter GL, Robboy S, et al. The 16. Leitao MM Jr, Han G, Lee LX, et al.
rate of concurrent EC at time of hyster- molecular genetics and morphometry-based Complex atypical hyperplasia of the uterus:
ectomy is slightly higher in this cohort endometrial intraepithelial neoplasia classifica- characteristics and prediction of underlying
(48.5%) than in the prospective Gyne- tion system predicts disease progression in carcinoma risk. Am J Obstet Gynecol
endometrial hyperplasia more accurately than 2010;203:349.
cologic Oncology Group 167 cohort, the 1994 World Health Organization classifica- 17. Costales AB, Schmeler KM, Broaddus R,
which demonstrated a concurrent EC tion system. Cancer 2005;103:2304–12. Soliman PT, Westin SN, Ramirez PT, et al.
risk of 43%.4 This may be explained by 4. Trimble CL, Kauderer J, Zaino R, et al. Clinically significant endometrial cancer risk
the fact that we are a tertiary referral Concurrent endometrial carcinoma in women following a diagnosis of complex atypical hy-
center and thus receive referrals from with a biopsy diagnosis of atypical endome- perplasia. Gynecol Oncol 2014;135:451–4.
trial hyperplasia: a Gynecologic Oncology 18. Kurman RJ, Norris HJ. Evaluation of criteria
many different centers with varying de- Group study. Cancer 2006;106:812–9. for distinguishing atypical endometrial hyper-
grees of pathologist expertise. Although 5. Parazzini F, La Vecchia C, Bocciolone L, plasia from well-differentiated carcinoma. Can-
the preoperative diagnosis was not cen- Franceschi S. The epidemiology of endometrial cer 1982;49:2547–59.
trally reviewed for all cases, this does cancer. Gynecol Oncol 1991;41:1–16. 19. Creasman WT, Morrow CP, Bundy BN,
reflect the “real world” of gynecologic 6. Parazzini F, La Vecchia C, Negri E, Fedele L, Homesley HD, Graham JE, Heller PB. Surgical
Balotta F. Reproductive factors and risk of pathologic spread patterns of endometrial can-
care. Strengths of this study include the cer. A Gynecologic Oncology Group Study.
endometrial cancer. Am J Obstet Gynecol
large cohort, final pathology review at a 1991;164:522–7. Cancer 1987;60(8 Suppl):2035–41.
single institution, and inclusion of 7. Clark TJ, Mann CH, Shah N, Khan KS, 20. Randall ME, Filiaci VL, Muss H, et al. Ran-
objective variables to better predict the Song F, Gupta JK. Accuracy of outpatient domized phase III trial of whole-abdominal irra-
risk of underlying cancer at the time of endometrial biopsy in the diagnosis of endo- diation versus doxorubicin and cisplatin
metrial cancer: a systematic quantitative review. chemotherapy in advanced endometrial carci-
hysterectomy for EIN. noma: a Gynecologic Oncology Group study.
BJOG 2002;109:313–21.
8. Bedner R, Rzepka-Gorska I. Hysteroscopy J Clin Oncol 2006;24:36–44.
Conclusion with directed biopsy versus dilatation and 21. May K, Bryant A, Dickinson HO, Kehoe S,
Ultimately, the decision to refer a pa- curettage for the diagnosis of endometrial hy- Morrison J. Lymphadenectomy for the man-
tient to gynecologic oncology is perplasia and cancer in perimenopausal agement of endometrial cancer. Cochrane
dependent on a discussion of risks and women. Eur J Gynaecol Oncol 2007;28:400–2. Database Syst Rev 2010;1:CD007585.
9. American College of Obstetricians and 22. Mariani A, Dowdy SC, Cliby WA, et al.
benefits between the patient and her Prospective assessment of lymphatic
Gynecologists. Committee Opinion Number
provider. Patients diagnosed with EIN 631: Endometrial intraepithelial neoplasia. dissemination in endometrial cancer: a para-
and an EMS of <2 cm should be 2015. Available at: https://www.acog.org/Clinical- digm shift in surgical staging. Gynecol Oncol
counseled on their underlying risk of Guidance-and-Publications/Committee-Opinions/ 2008;109:11–8.
carcinoma and potential lymph node Committee-on-Gynecologic-Practice/Endometrial- 23. Jewell EL, Huang JJ, Abu-Rustum NR,
Intraepithelial-Neoplasia. Accessed March 1, Gardner GJ, Brown CL, Sonoda Y, et al.
involvement. Most importantly, our Detection of sentinel lymph nodes in minimally
2019.
results suggest that all patients with an 10. Trimble CL, Method M, Leitao M, et al. invasive surgery using indocyanine green and
EMS of 2 cm should be considered Management of endometrial precancers. Obstet near-infrared fluorescence imaging for uterine
for referral to a gynecologic oncologist, Gynecol 2012;120:1160–75. and cervical malignancies. Gynecol Oncol
given the high odds of underlying 11. Karamursel BS, Guven S, Tulunay G, 2014;133:274–7.
Kucukali T, Ayhan A. Which surgical proced- 24. Tanner EJ, Sinno AK, Stone RL,
endometrial cancer. Because almost
ure for patients with atypical endometrial hy- Levinson KL, Long KC, Fader AN. Factors
half of the patients with an EMS of 2 perplasia? Int J Gynecol Cancer 2005;15: associated with successful bilateral sentinel
cm will meet criteria for lymph node 127–31. lymph node mapping in endometrial cancer.
assessment, strong consideration 12. Whyte JS, Gurney EP, Curtin JP, Blank SV. Gynecol Oncol 2015;138:542–7.
should be given to SLN dissection at Lymph node dissection in the surgical man- 25. National Comprehensive Cancer Network.
agement of atypical endometrial hyperplasia. Uterine neoplasms (Version 3.2017) 2017. Avail-
the time of hysterectomy for EIN. This
Am J Obstet Gynecol 2010;202:176. able at: https://www.nccn.org/professionals/
will provide the patient with an 13. Touhami O, Gregoire J, Renaud MC, physician_gls/pdf/uterine.pdf. Accessed March
adequate risk assessment (staging) in Sebastianelli A, Grondin K, Plante M. The utility 18, 2019.
case of cancer diagnosis on final pa- of sentinel lymph node mapping in the man- 26. Holloway RW, Abu-Rustum NR, Backes FJ,
thology, and would avoid a possible agement of endometrial atypical hyperplasia. et al. Sentinel lymph node mapping and staging
second surgery and/or complete Gynecol Oncol 2018;148:485–90. in endometrial cancer: a Society of Gynecologic
14. Rossi EC. Complex atypical hyperplasia: Oncology literature review with consensus rec-
lymphadenectomy. n when is it appropriate to refer? ObGyn ommendations. Gynecol Oncol 2017;146:
News. 2017. Available at: https://www. 405–15.
mdedge.com/obgyn/article/154533/gynecologic- 27. Barlin JN, Khoury-Collado F, Kim CH, et al.
References cancer/complex-atypical-hyperplasia-when-it- The importance of applying a sentinel lymph
1. Sherman ME. Theories of endometrial carci- appropriate-refer. Accessed March 18, 2019. node mapping algorithm in endometrial cancer
nogenesis: a multidisciplinary approach. Mod- 15. Suh-Burgmann E, Hung YY, staging: beyond removal of blue nodes. Gynecol
ern Pathol 2000;13:295–308. Armstrong MA. Complex atypical endometrial Oncol 2012;125:531–5.

JANUARY 2020 American Journal of Obstetrics & Gynecology 60.e6


Original Research GYNECOLOGY ajog.org

28. Nout RA, Smit VT, Putter H, et al. Vaginal randomised, phase 3 trial. Lancet Oncol Biostatistics (Mr Benedict and Dr Hade), Department of
brachytherapy versus pelvic external beam 2018;19:295–309. Biomedical Informatics, The Ohio State University College
radiotherapy for patients with endometrial can- of Medicine, Columbus, OH.
cer of high-intermediate risk (PORTEC-2): an Author and article information Received March 19, 2019; revised July 5, 2019;
open-label, non-inferiority, randomised trial. From the Division of Gynecologic Oncology (Drs Vetter, accepted Aug. 3, 2019.
Lancet 2010;375:816–23. Bixel, Copeland, Cohn, Fowler, O’Malley, Salani, and The authors report no conflicts of interest.
29. de Boer SM, Powell ME, Mileshkin L, et al. Backes), Department of Obstetrics/Gynecology, The Ohio These findings were presented in part at the 2017
Adjuvant chemoradiotherapy versus radio- State University College of Medicine, Columbus, OH; Di- Society of Gynecologic Oncology Winter Meeting in
therapy alone for women with high-risk endo- vision of Gynecologic Oncology (Dr Smith), Department of Breckenridge, CO, Jan. 26e28, 2017.
metrial cancer (PORTEC-3): final results of an Obstetrics/Gynecology, University of MissourieKansas Corresponding author: Floor Backes, MD. Floor.
international, open-label, multicentre, City School of Medicine, Kansas City, MO; Center for backes@osumc.edu

60.e7 American Journal of Obstetrics & Gynecology JANUARY 2020