Professional Documents
Culture Documents
Endometrial Intraepithelial Neoplasia
Endometrial Intraepithelial Neoplasia
COMMITTEE OPINION
Number 631 • May 2015
(Reaffirmed 2019)
Conclusions and Recommendations based upon evidence that has become available
Sensitive and accurate diagnosis of true premalignant since the creation of the more widely used WHO94
endometrial lesions can reduce the likelihood of devel- schema (in which atypical hyperplasia is equated
oping invasive endometrial cancer. Based on avail- with precancerous behavior). The preferred termi-
able data and expert opinion, the American College nology is “endometrial intraepithelial neoplasia”
of Obstetricians and Gynecologists and the Society of (rather than “atypical endometrial hyperplasia”).
Gynecologic Oncology make the following consensus • Regarding tissue sampling, hysteroscopy, while not
recommendations: required, is recommended with directed dilation
• The endometrial intraepithelial neoplasia schema and curettage (D&C) to include any discrete lesions
seems to be preferable to the 1994 four-class World as well as the background endometrium. This will
Health Organization (WHO94) schema. Pathologic provide the best opportunity to confirm the diag-
diagnosis of premalignant lesions should use criteria nosis of a true premalignant endometrial lesion and
and terminology that clearly distinguish between exclude an associated endometrial carcinoma. When
clinicopathologic entities that are managed differ- clinically appropriate, total hysterectomy for endo-
ently. At present, the endometrial intraepithelial metrial intraepithelial neoplasia provides definitive
neoplasia schema is tailored most closely to this assessment of a possible concurrent carcinoma and
objective, incorporating modified pathologic criteria effectively treats premalignant lesions.
• Supracervical hysterectomy, morcellation, and endo- precancerous behavior). “Endometrial intraepithelial
metrial ablation are unacceptable for treatment of neoplasia” (rather than “atypical endometrial hyper-
endometrial intraepithelial neoplasia. plasia”) is the preferred terminology that will be used
• Systemic or local progestin therapy is an unproven throughout this document.
but commonly used alternative to hysterectomy Endometrial Hyperplasia Classification
that may be appropriate for women who are poor
Systems
surgical candidates or who desire to retain fertility.
There are currently two systems of endometrial precancer
• Posthormonal treatment surveillance after nonsurgi-
nomenclature in common usage: 1) the WHO94 schema
cal management of endometrial intraepithelial neo-
and 2) the endometrial intraepithelial neoplasia diagnos-
plasia may include serial endometrial sampling every
tic schema developed by the International Endometrial
3–6 months, but the appropriate frequency has not
Collaborative Group (2). The WHO94 schema classifies
yet been determined.
histology based on glandular complexity and nuclear
atypia and is comprised of four categories of risk clas-
Background sification: 1) simple hyperplasia, 2) complex hyperpla-
Endometrial hyperplasia is of clinical significance because sia, 3) simple hyperplasia with atypia, and 4) complex
it is often a precursor lesion to adenocarcinoma of the hyperplasia with atypia. These categories are descriptive
endometrium (1, 2). The precursor lesion of type I endo- in nature, and interpretation is subjective; accordingly,
metrioid adenocarcinoma is endometrial intraepithelial studies indicate poor reproducibility of the individual
neoplasia. Estrogenic stimulation of the endometrium, case classification (3, 4). Moreover, the individual catego-
unopposed by progestins, causes proliferative glandu- ries do not suggest specific management algorithms. This
lar epithelial changes. This finding, due to prolonged older schema is the one most commonly used by patholo-
hormonal exposure, is biologically distinct from true gists, but transitioning to the endometrial intraepithelial
precancerous lesions and true neoplasia. Making the neoplasia nomenclature would provide greater benefit to
distinction between hyperplasia and true precancerous clinical management.
lesions or true neoplasia has significant clinical effect In the endometrial intraepithelial neoplasia schema,
because their differing cancer risks must be matched endometrial precancer is termed “endometrial intraepi-
with an appropriate intervention to avoid undertreat- thelial neoplasia” (5, 6). Pathologic criteria were used
ment or overtreatment. The focus of this Committee to develop three disease categories: 1) benign (benign
Opinion is the classification of endometrial hyperplasia endometrial hyperplasia), 2) premalignant (endometrial
and treatment options. Gynecologists should be aware of intraepithelial neoplasia), and 3) malignant (endometrial
the two nomenclature schemas and that the endometrial adenocarcinoma, endometrioid type, well differentiated)
intraepithelial neoplasia schema seems to be preferable (Table 1 and Table 2). By applying the endometrial intra-
to the WHO94 schema. Pathologic diagnosis of prema- epithelial neoplasia schema to routinely obtained endo-
lignant lesions should use criteria and terminology that metrial tissues, pathologists present the clinician with
clearly distinguish between clinicopathologic entities a disease-specific classification that informs treatment
that are managed differently. At present, the endometrial decisions. Diagnosis using the endometrial intraepithelial
intraepithelial neoplasia schema is tailored most closely neoplasia schema has been confirmed as prognostic in
to this objective, incorporating modified pathologic cri- several retrospective studies and one prospective study
teria based upon evidence that has become available (7–9). Two of these studies also suggest that interobserver
since the creation of the more widely used WHO94 reproducibility using the endometrial intraepithelial
schema (in which atypical hyperplasia is equated with neoplasia schema can be greater than with the WHO94
Benign endometrial hyperplasia Diffuse Prolonged estrogen effect Hormonal therapy, symptomatic
Endometrial intraepithelial neoplasia Focal progressing to diffuse Precancerous Hormonal therapy or surgery
Endometrial adenocarcinoma, Focal progressing to diffuse Malignant Surgery, stage based
endometrioid type, well
differentiated
Architecture Area of glands greater than stroma (volume percentage stroma less than 55%)
Cytology Cytology differs between architecturally crowded focus and background
Size greater than 1 mm Maximum linear dimension exceeds 1 mm
Exclude mimics Benign conditions with overlapping criteria (ie, basalis, secretory, polyps, repair)
Exclude cancer Carcinoma if maze-like glands, solid areas, or appreciable cribriforming
schema (7, 9), which is why gynecologic oncologists prefer the best opportunity to confirm the diagnosis of a true
the endometrial intraepithelial neoplasia schema. premalignant endometrial lesion and exclude an associ-
ated endometrial carcinoma. The small volume of tissue
Precancer Diagnosis: Endometrial Sampling obtained by currently available technologies for sampling
and Imaging the endometrium may limit an accurate assessment of
Sensitive and specific detection of endometrial precancer cancer risk. Current diagnostic schema should include an
and exclusion of coexisting carcinoma are prerequi- assessment of sample adequacy, as is recommended for
sites for management of patients with premalignant evaluation of cervical cytology specimens (15).
endometrial lesions. Excluding concurrent carcinoma
by endometrial suction curette is especially problematic: Diagnosis of Endometrial Cancer Among
approximately 40% of patients who receive a premalig- Women With Postmenopausal Bleeding
nant endometrial intraepithelial neoplasia diagnosis by Transvaginal ultrasonography has excellent negative pre-
endometrial suction curette receive a carcinoma diagno- dictive value for endometrial cancer in women with
sis by using a hysterectomy specimen (8, 10). postmenopausal bleeding. When transvaginal ultraso-
The accuracy of D&C compared with endometrial nography is performed for patients with postmenopausal
suction curette in diagnosing precancer and exclud- bleeding and an endometrial thickness of 4 mm or less is
ing concurrent carcinoma is unclear. Both have sam- found, endometrial sampling is not required because of
pling limitations: approximately 60% of D&C specimens the very low risk of uterine malignancy in these patients
sample less than one half of the uterine cavity (11). The (16). An endometrial thickness greater than 4 mm in
method of sampling is less important if management a patient with postmenopausal bleeding should trig-
includes definitive treatment with a hysterectomy, which ger alternative evaluation (such as sonohysterography,
eliminates the risk of failure to diagnose an endometrial office hysteroscopy, or endometrial biopsy), as should an
cancer. Dilation and curettage and endometrial suction inability to adequately visualize endometrial thickness.
curette sampling devices have been reported to yield The significance of an endometrial thickness greater than
equal rates of cancer detection in patients with abnormal 4 mm in an asymptomatic, postmenopausal patient has
uterine bleeding (12). A single-institution retrospective not been established, and this finding need not routinely
series found that D&C used to diagnose endometrial trigger evaluation (16). The utility of ultrasonographic
intraepithelial neoplasia was less likely to miss cancer depiction of endometrial thickness for ruling out malig-
(which was evident on subsequent hysterectomy) than nancy is limited to the postmenopausal patient who has
the use of endometrial suction curette (27% compared bleeding.
with 46%, respectively) (13). Mass lesions that impinge
upon the uterine cavity may deflect flexible endome- Management of Endometrial Intraepithelial
trial suction curette devices, which prevents adequate Neoplasia
assessment of the endometrial cavity. Hysteroscopy with The primary objectives in a patient in whom endometrial
directed biopsy is more sensitive than D&C in the diag- intraepithelial neoplasia has been newly diagnosed are
nosis of uterine lesions (14). Regarding tissue sampling, the following: ruling out a concurrent adenocarcinoma,
hysteroscopy, while not required, is recommended with designing a treatment plan that can accommodate delayed
directed D&C to include any discrete lesions as well discovery of an occult carcinoma, and preventing the
as the background endometrium. This will provide progression to endometrial cancer. Total hysterectomy
per month) or micronized progesterone in vaginal cream management. Posthormonal treatment surveillance after
(100 mg for 12–14 days per month) when treated for nonsurgical management of endometrial intraepithelial
3 months (31–34). However, if endometrial intraepithe- neoplasia may include serial endometrial sampling every
lial neoplasia is present, there is a higher incidence of 3–6 months, but the appropriate frequency has not yet
failure of medical management and subsequent develop- been determined.
ment of cancer (33). There is no consensus on the preferred nonsurgi-
Systemic or local progestin therapy is an unproven cal treatment of endometrial intraepithelial neoplasia;
but commonly used alternative to hysterectomy that therefore, it is difficult to recommend a standard regi-
may be appropriate for women who are poor surgical men. Several proposed treatment strategies are shown in
candidates or who desire to retain fertility. In addi- Table 3. Treatment with an oral progestin or a 5-year
tion to systemic administration of hormonal agents, levonorgestrel IUD is a reasonable first option and, based
some studies have investigated the use of intrauterine on the patient’s clinical situation (eg, no longer desires
devices (IUDs) for the delivery of progestins. The 5-year fertility, has completed childbearing, or has become an
levonorgestrel-releasing intrauterine system (levonorg- acceptable-risk surgical candidate), should be continued
estrel IUD) provides a potential alternative to oral pro- for 12 months or more unless progression is identified. For
gestogen. Local-acting progesterone has an effect on the many women, the underlying hormonal cause of endo-
endometrium that is several times stronger than that metrial intraepithelial neoplasia remains after therapy is
exerted by systemic products and has a decreased sys- completed. Sloughing of the target lesion may be followed
temic effect. Most studies have a relatively small sample by recurrence if treatment is not continued indefinitely.
size, but in a study of 105 women with simple, complex, Obesity is associated with an increased incidence of
or atypical hyperplasia, regression of hyperplasia with endometrial cancer. Because endometrial intraepithelial
use of a levonorgestrel IUD was up to 90%, although neoplasia is often an antecedent of endometrial cancer,
only approximately 67% in the presence of atypia (28). clinicians may counsel patients about weight loss or bar-
A systematic review and meta-analysis found a pooled iatric surgery to reduce the risk of recurrence. Long-term
regression rate of 69% (95% confidence interval, 58–83) systemic medical treatment to prevent reappearance of
in 14 studies (n=189) of women with atypical hyperplasia endometrial intraepithelial neoplasia requires awareness
treated with oral progestins. Pooling the seven studies of potential adverse effects. Edema, gastrointestinal dis-
(n=36) of women with atypical hyperplasia treated with turbances, and thromboembolic events are infrequent
the levonorgestrel IUD found a regression rate of 90% with these treatments, thereby making medical manage-
(95% confidence interval, 62–100) (35). ment a reasonable therapeutic option for patients for
There are still several unresolved issues regarding whom surgical management is not optimal (36).
hormonal treatment of endometrial intraepithelial neo- References
plasia. The optimal treatment dose and duration has not
1. Sherman ME. Theories of endometrial carcinogenesis: a
been determined, nor has it been determined whether multidisciplinary approach. Mod Pathol 2000;13:295–308.
treatment should be cyclic or continuous. The appropri- [PubMed] [Full Text] ^
ate length of follow-up after treatment also has not been
2. Silverberg SG, Kurman RJ, Nogales F, Mutter GL, Kubik-
defined clearly. Appropriate measures of the clinical and Huch RA, Tavassoli FA. Epithelial tumours and related
histologic response to progestogen treatment also are lesions. In: Tavassoli FA, Devilee P, editors. Pathology and
lacking. Full examination of the endometrium is required genetics of tumours of the breast and female genital organs.
to measure regression, persistence, or progression of endo- World Health Organization classification of tumours. Lyon
metrial intraepithelial neoplasia. Examination of the (France): IARC Press; 2003. p. 221–32. ^
entire uterus after hysterectomy is considered ideal but 3. Kendall BS, Ronnett BM, Isacson C, Cho KR, Hedrick L,
is not an option for patients who receive nonsurgical Diener-West M, et al. Reproducibility of the diagnosis of