Investigation of Abnormal Uterine Bleeding in Peri and

Document Sample
Investigation of Abnormal Uterine Bleeding in Peri and Powered By Docstoc

Investigation of Abnormal Uterine Bleeding
in Peri- and Postmenopausal Women
Linda D. Bradley, MD
Malgorzata E. Skaznik-Wikiel, MD

The ovarian life cycle spans 4 decades, with most women having                        age 50.8 years will experience men-
                                                                                      strual abnormalities.1
approximately 450–500 menstrual cycles during a lifetime. Irregularity
                                                                                         Anovulation is the most common
in menstrual cycles is frequent during the perimenopausal years, but                  culprit in menstrual dysfunction dur-
is often only a nuisance, with malignancy rarely the cause. Such                      ing the transition years. Rarely should
bleeding becomes greater cause for concern when it occurs during                      surgical intervention be the first step
                                                                                      in the evaluation of perimenopausal
the postmenopausal years.                                                             abnormal bleeding. Rather, the gy-
   Each year there are more than 2.5 million visits to gynecologists for              necologist should evaluate the most
the evaluation of abnormal uterine bleeding. Physicians must maintain                 likely etiology via medical history and
                                                                                      pelvic evaluation and, in the absence
a low threshold for endometrial assessment in postmenopausal women                    of any suggestion of pathology, should
who present with abnormal bleeding. Accurately determining the                        first aggressively treat medically with
etiology of the bleeding permits appropriate treatment, minimizes                     hormonal contraceptives, proges-
                                                                                      terone therapy or with a medicated
unnecessary delays in therapy and prevents needless worry in
                                                                                      intrauterine system (Mirena intra-
patients. Fortunately, medical therapy is effective and appropriate for               uterine device).
the majority of patients; only when response to medical therapy is less                  In addition to anovulatory cycles,
than satisfactory should organic disease, intrauterine pathology and                  reproductive-age patients may also
                                                                                      have intrauterine structural abnor-
endocrine abnormalities be ruled out.                                                 malities that contribute to abnormal
   This article will address the salient issues and office-based options              bleeding. The most conservative ap-
available for the evaluation of abnormal uterine bleeding in this                     proaches to management of abnor-
                                                                                      mal bleeding in such cases include
group of women.                                                                       “watchful waiting”, medical therapy
                                                                                      and, when such therapy fails or the
Abnormal Pre- and                          related complications, disorders of he-    patient is markedly anemic, visuali-
Perimenopausal Bleeding                    mostasis, trauma and infection.            zation or imaging of the endometrial
Abnormal uterine bleeding in repro-           Menstrual aberrations occur at the      cavity. Treatment initiatives include
ductive-age women is common, lead-         extremes of the reproductive life cycle.   methods to regulate menstrual cycles,
ing to one-third of outpatient visits      Within the first 1–5 years after the       minimize blood loss and prevent
by this population.The most frequent       onset of menstruation, and also be-        complications of chronic unopposed
culprits in this age group include         ginning by the late 30s and 40s,           estrogen stimulation.
anovulation, polycystic ovary syn-         periodic or persistent menstrual dys-         Seltzer et al2 retrospectively followed
drome, structural abnormalities (pol-      function is typical. Approximately         500 perimenopausal women and sys-
yps, fibroids), endometrial hyperplasia,   50% of women by age 45.5 years,            tematically noted that the classic
cancer, foreign bodies, pregnancy-         75% by age 47.8 years and 95% by           categories of bleeding typical of the

                                                                                              NOVEMBER /DECEMBER 2008         13

perimenopause included hypomenor-         were detected only with SIS. The in-      Postmenopausal Bleeding
rhea (70%); menorrhagia, metrorrha-       vestigators found that SIS provided       The average age at which women ex-
gia and/or hypermenorrhea (18%);          more accurate information regard-         perience menopause has remained
and sudden cessation of menses (12%).     ing the presence of intracavitary         steady, occurring in 80% of women
Using multiple hormonal analyses,         pathology.                                by age 51, and 95% of women by age
Landgren et al3 evaluated 13 women
longitudinally, between 4 and 9 years
before menopause. They noted in-
creased frequency of anovulatory cycles
characterized by elevated follicle-
stimulating hormone (FSH) concen-
trations 30 cycles before final men-
struation occurred. Women with
hormonal evidence of ovulatory cycles
did not have FSH levels that rose

Evaluation (Perimenopausal
Transvaginal ultrasound (TVUS) has
been the gold standard of imaging
in the postmenopausal patient; how-
ever, in the reproductive-age patient
TVUS is less accurate in the diagno-
sis of focal endometrial pathology.4
The sensitivity of TVUS for detecting
                                          Numerically, the highest rate of PMB occurs within the
all intracavitary lesions has ranged      5 years after spontaneous amenorrhea, and gradually
from 48% to 96% and the specificity
                                          declines thereafter.
has ranged from 68% to 95%.5,6 A
recent study by Breitkopf and col-
leagues7 highlights the limitations          In almost all studies to date, SIS     55, with virtually all women experi-
of TVUS alone in reproductive-age         has been shown to be more sensitive       encing cessation of menses by age 58.
patients and, when possible, strongly     and reliable than TVUS in detecting       Postmenopausal bleeding (PMB) is
advocates the addition of saline          endometrial polyps and submucosal         defined as any bleeding that occurs
infusion to improve detection of in-      fibroids. When used without con-          12 months after the last menstrual
tracavitary lesions and myometrial        comitant endometrial biopsy, SIS is       period. After years of amenorrhea,
abnormalities in this age group.          less sensitive in diagnosing endome-      the onset of any vaginal bleeding is
When saline infusion sonography           trial hyperplasia in the perimenopausal   cause for concern on the part of both
(SIS) is not available, in-office hys-    population.8                              clinician and patient. Even small
teroscopy can be used to quickly             In-office hysteroscopy performed       amounts of bleeding should elicit
and comfortably evaluate the en-          for a more comprehensive evaluation       prompt and thorough evaluation.
dometrium. In this study the investi-     of the endometrium is a reasonable        Additionally, acyclical vaginal bleed-
gators evaluated 206 premenopausal        option when a patient continues to        ing in postmenopausal women receiv-
women who had normal endometrial          have abnormal uterine bleeding de-        ing hormone therapy (HT) warrants
echoes. They determined that one in       spite medical therapy, and has under-     evaluation.
six lesions was missed when TVUS          gone only TVUS (without SIS). The            Numerically, the highest rate of
was used alone. Specifically, of 80       Breitkopf study suggests that an en-      PMB occurs within the 5 years after
subjects with an endometrial echo of      dometrial echo of ≤ 5 mm cannot           spontaneous amenorrhea, and gradu-
≤5 mm, 11 had endometrial polyps          exclude benign endometrial pathol-        ally declines thereafter. In a prospective,
and 5 had submucosal fibroids that        ogy in premenopausal patients.            observational, population-based study,


297 postmenopausal women completed        more common in white women, black           usually concerned about having missed
1 year of daily recording of bleeding.9   women are twice as likely to die from       a focal lesion. Karlsson et al reviewed
PMB occurred in 409 per 1,000 per-        the disease; black women tend to be         1,168 cases of PMB evaluated with
son-years in the first year immediately   diagnosed later, have a more advanced       TVUS followed by dilation and curet-
after menopause. When more than 3         stage of cancer upon diagnosis, and a       tage,13 including 351 women (ages 41
years elapsed, the rate plummeted to      more lethal histologic subtype (serous      to 91 years) taking HT. They found
42 per 1,000 person-years.                tumors) than do white women.                that PMB was due to the following
   Endometrial cancer. Endometrial           Clinicians must carefully evaluate       causes: endometrial atrophy (>59% of
cancer is the most common genital         any abnormal bleeding in older              cases); polyps (12%); endometrial can-
female malignancy,10 accounting for       women, since 95% of endometrial             cer (10%); endometrial hyperplasia
nearly half of all new cases of genital   cancers occur in women age 40 and           (9.8%); hormonal effect (7%); hydro-
cancer. The probability of endome-        older, and because endometrial hyper-       metra, pyometra and hematometra
trial cancer in women presenting with     plasia, the precursor state, may pre-       (2%); and cervical cancer (<1%).
PMB is approximately 5–15%,10 and         cede the diagnosis.11                          TVUS endometrial measurement
the lifetime risk of being diagnosed         Ovarian, cervical and fallopian tube     (endometrial echo/endometrial stripe)
with the disease is 2.7%.10 The median    cancers. Gredmark and colleagues re-        is helpful in categorizing patients into
age at diagnosis is 63 years. Risk fac-   port that, together, ovarian and cervi-     a low- versus high-risk group. The
tors include obesity, use of unopposed    cal cancer occur in 3% of women with        exact cutoff measurement chosen is a
estrogen therapy, prolonged tamox-        PMB.12 Fallopian tube carcinoma, a          function of the sensitivity and speci-
ifen use, anovulation, polycystic ovary   true rarity, must also be considered        ficity sought. Most clinicians use a
syndrome, estrogen-secreting tumors,      when persistent vaginal bleeding            cutoff of 5 mm to define a low-risk
nulliparity, late menopause and history   occurs despite a negative evaluation of     patient group, whose combined risk
of complex endometrial hyperplasia        the endometrium. These statistics are       for cancer and atypical hyperplasia
with atypia.                              important to remember, especially           ranges from 2% to 3%. The initial
   Although the total mortality rate      when a patient continues to bleed af-       goal of TVUS was not to replace
for women with endometrial cancer         ter a negative evaluation of the uterine    endometrial biopsy, but to decrease
has decreased since the 1980s, this is    cavity. The adnexa and the bladder          the number of endometrial biopsies
not an indolent disease. Stage for        should be evaluated (urine cytology         needed.14 Ultrasonographers painstak-
stage, endometrial cancer is just as      and bladder imaging) when endome-           ingly sought to find the specific endo-
lethal as ovarian cancer. But because     trial evaluation is unremarkable in the     metrial echo length for which the
of a usually excellent prognosis, endo-   presence of genital bleeding.               likelihood of missing endometrial
metrial cancer is often referred to as                                                cancer would be exceedingly low.15
the “curable cancer” because of early     Evaluation of Postmenopausal                   Criteria for stratification into low-
detection and prompt surgical inter-      Bleeding                                    risk groups include age less than 70
vention. Unlike women with ovarian        The work-up of patients with PMB            years, multiparous status, bleeding that
cancer, most patients with endome-        should always be thorough. Clinical         occurs within 1 year of menopause
trial cancer are symptomatic and pres-    evaluation of PMB can be accom-             and absence of diabetes. In contrast,
ent with PMB or a lengthy bout of         plished with a number of diagnostic         patients with an endometrial echo >5
perimenopausal bleeding. Indeed, the      methods and tools.                          mm have an increased risk of en-
presence of symptoms (bleeding, pelvic       Endometrial biopsy and TVUS. Tra-        dometrial cancer and atypical hyper-
cramping or unexplained chronic vagi-     ditionally, endometrial biopsy has          plasia approaching 5% or more.
nal discharge) is the tangible hallmark   been considered the gold standard           Patients with all four risk factors—
of endometrial cancer, which should       for the evaluation of PMB, with both        diabetes, nulliparity, age >70 years and
prompt early investigation, identifi-     clinician and patient reassured by a        bleeding that occurs more than 1 year
cation and initiation of treatment,       normal biopsy result. The challenge,        after menopause—have a relative risk
assuring a better prognosis. Overall      however, is the two-thirds of patients      of 1.8 compared with a low-risk pop-
5-year survival is 86%; when the dis-     with PMB who have an endometrial            ulation.16 However, many evaluations
ease is confined within the uterus 5-     biopsy that demonstrates endometrial        of the usefulness of the endometrial
year survival rates peak at 97%.11        atrophy or “tissue insufficient for di-     measurement have included asympto-
Although endometrial cancer is 40%        agnosis.” In these cases the clinician is   matic patients. It may therefore be

                                                                                             NOVEMBER /DECEMBER 2008        15

that patients presenting with bleeding     the PMB work-up to exclude cervical          malignant lesion of the endocervix or
have a higher detection rate than          or endocervical causes of bleeding. A        endometrium in 10–40% of cases.19
asymptomatic patients.                     liquid-based Pap smear is warranted          Therefore, all women over age 35 and
   SIS and hysteroscopy. TVUS and,         for all women with new-onset PMB,            younger women with unexplained
when available, SIS should be con-         even when the patient has undergone          bleeding and atypical glandular cells
sidered in the work-up. When the           cervical cytology within the past 1 to       on the smear require further, more in-
patient’s bleeding persists despite a      3 years. Even though newer cervical          vasive testing.19
negative biopsy or normal TVUS,            cytology guidelines propose less fre-
continue to evaluate with additional       quent Pap tests after age 60, there is       Rethinking Evaluation Tools for
technology. If only TVUS (Figures 1        a bimodal distribution of cervical           Peri- and Postmenopausal
and 2) and endometrial biopsy were         cancer, as well as its indolent endo-        Bleeding
initially utilized, direct inspection of   cervical adenocarcinoma counterpart.         Do we need to rethink the role of
the endometrium with in-office hys-        A smear with atypical glandular cells        endometrial biopsy in the evaluation of
teroscopy or SIS should be considered      is associated with a premalignant or         patients with abnormal uterine bleed-
if bleeding recurs. The interface cre-
ated by fluid during SIS effectively
delineates intraluminal defects (Fig-
ures 3 and 4). Likewise, direct visual-
ization of the endometrium with
hysteroscopy detects intracavitary
pathology.17 Several authors have
found that when women were fol-
lowed after a negative initial endome-
trial biopsy, 20% with persistent or
episodic vaginal bleeding were ulti-
mately diagnosed with either com-
plex atypical hyperplasia or cancer.18
Vigilance and the use of additional
technology are important to minimize           Figure 1. Transvaginal ultrasound, longitudinal view. Uterine measurement:
the chance of missing intracavitary            8.2 cm x 4.8 cm x 5.4 cm. The endometrial echo is ill-defined, with a
pathology.                                     hypoechoic mass.
   Pelvic examination. Pelvic examina-
tion should include thorough inspec-
tion of the vagina, vaginal fornices,
vulva, vaginal vault and urethra for
lesions. Vaginal atrophy is a frequent
cause of abnormal bleeding; thin, fri-
able tissue that bleeds easily with in-
tercourse or spontaneously is common
with each advancing decade. Any vul-
var lesion with chronic excoriation
should be biopsied liberally. Vagin-
oscopy can be used to detect subtle
vaginal ulceration, foreign bodies and
lesions. When vaginal atrophy is
noted, an excellent response to oral or
topical estrogen therapy is the norm.
                                               Figure 2. Transvaginal ultrasound, coronal view. Uterine measurement: 8.2 cm
   Cervical cytology. While not a
                                               x 4.8 cm x 5.4 cm. The endometrial echo is ill-defined, with a hypoechoic mass.
screening tool for endometrial cancer,
the Pap test must be included in


ing? Is endometrial biopsy, when used         harbor endometrial cancer when the        hyperplastic endometrium. For this
alone, an outdated modality in the al-        endometrial echo is < 5 mm and the        reason, the clinician should not end
gorithm of the evaluation of abnormal         patient has minimal risk factors.21       the work-up in a patient who contin-
bleeding? Most studies demonstrate            The adnexa should also be imaged to       ues to have symptoms but has an
the inadequacy of blind endometrial           rule out adnexal pathology. Using a       endometrial echo < 5 mm. When
sampling, including dilatation and            threshold of < 5 mm, an endometrial       symptoms persist, or the amount of
curettage, in women with focal lesions.20     echo cutoff is the most cost-effective    bleeding is profuse, then reevaluation
   Increasingly, TVUS is proposed as          diagnostic strategy for detecting endo-   with direct visualization via hystero-
the first-line, minimally invasive tool       metrial cancer.22 However, an endo-       scopically directed biopsy or SIS is
in the evaluation of PMB. In general,         metrial echo < 5 mm can harbor other      indicated. When a focal lesion is de-
if TVUS is used alone, an endometrial         nonmalignant pathology that causes        tected, hysteroscopic resection or di-
stripe that is well delineated in its en-     PMB, including endometrial atrophy,       rected biopsy is warranted.
tirety, and is homogeneous, without           polyps, endometritis, submucosal fi-         Similarly, if a patient with PMB
fluid or irregularities will unlikely         broids, pyometra, and proliferative and   has a thickened endometrial echo,
                                                                                        especially > 5 mm, and has endome-
                                                                                        trial curettings obtained blindly or
                                                                                        biopsy that is insufficient, additional
                                                                                        evaluation with hysteroscopy or SIS is
                                                                                        mandatory. Most of these patients will
                                                                                        have a focally growing lesion, which
                                                                                        can be hysteroscopically resected. Be-
                                                                                        nign pathology, such as polyps and
                                                                                        hyperplasia, causes bleeding. Less than
                                                                                        2% of polyps contain a malignancy or
                                                                                        hyperplasia with/without atypia. But
                                                                                        malignant changes within polyps can
                                                                                        rarely be detected with ultrasound
                                                                                        (including Doppler flow) or hystero-
                                                                                        scopic visualization alone, and even
    Figure 3. SIS, longitudinal view, demonstrating a 2.3 cm x 1.9 cm x 2.7 cm          small polyps < 2 cm can be malig-
    intracavitary submucosal fibroid with a right lateral wall attachment site.         nant.23 A polypoid growth can actu-
                                                                                        ally be a focal area of hyperplasia, a
                                                                                        malignancy or a mesenchymal tu-
                                                                                        mor.24 Obtaining histology is para-
                                                                                        mount when symptomatic abnormal
                                                                                        bleeding is encountered.
                                                                                           A thorough review of the litera-
                                                                                        ture clearly delineates that when the
                                                                                        endometrium is imaged entirely by
                                                                                        TVUS, and is homogeneous and <5
                                                                                        mm, it will unlikely harbor endome-
                                                                                        trial cancer. However, TVUS alone
                                                                                        rarely can determine the presence of
                                                                                        focal lesions. It is the benign focal le-
                                                                                        sions that usually cause episodic and
                                                                                        irregular bleeding. Facts to consider
                                                                                        include the following:
    Figure 4. SIS, coronal view, demonstrating a 2.3 cm x 1.9 cm x 2.7 cm
                                                                                        • Findings from a meta-analysis
    intracavitary submucosal fibroid with a right lateral wall attachment site.
                                                                                             of 35 studies involving 5,892
                                                                                             women and using an endometrial

                                                                                                NOVEMBER /DECEMBER 2008        17

     thickness cutoff of 5 mm25 re-        and should appear homogeneous and           65 articles that evaluated the role of
     vealed that this cutoff had > 92%     without fluid. If the endometrial echo      hysteroscopy in 26,345 women de-
     sensitivity for detecting benign      is indistinct, not visualized, irregular    termined that a positive hysteroscopy
     and malignant endometrial dis-        and has a heterogeneous appearance,         increased the probability of endome-
     ease (polyps, atypical hyperplasia    then additional surveillance with SIS       trial cancer to 71.8%, whereas a neg-
     or cancer) and 96% sensitivity        or hysteroscopy is mandatory. Addi-         ative result reduced the probability of
     for detecting endometrial can-
     cer. Therefore, a postmenopausal
     woman with a 10% pretest prob-        It is the benign focal lesions that usually cause episodic and
     ability of endometrial cancer has
     a 1% probability of cancer if her     irregular bleeding.
     TVUS has an endometrial echo
     < 5 mm.
•    Tsuda et al26 evaluated 600 post-     tionally, if bleeding recurs despite a      cancer to 0.6%. De Jong et al30 note
     menopausal women and used dif-        normal TVUS, then a direct view of          that the sensitivity and specificity for
     ferent endometrial echo cutoff        the endometrium is necessary. De-           the diagnosis of endometrial disease
     values based on the number of         spite a TVUS endometrial echo of < 4        was 78% and 95.8%, respectively. For a
     years since menopause. If the pa-     mm in patients with persistent bleed-       positive result the pretest probability
     tient was less than 5 years past      ing, either in-office hysteroscopy or       increased from 10.6% to 55.2%, and
     the onset of menopause the en-        SIS should be employed next to eval-        decreased to 2.8% with a negative re-
     dometrial echo was set at 4 mm;       uate for subtle intracavitary lesions.      sult. All women having hysteroscopy
     if more than 5 years menopausal          When the endometrial echo is > 4         should also have an endometrial biopsy,
     the cutoff was 3 mm. TVUS             mm, SIS or hysteroscopy should be           which provides a tissue diagnosis.
     demonstrated a 97.4% sensitivity,     performed. The distention of the en-            Hysteroscopy is more accurate in
     75.7% specificity, a 23.8% posi-      dometrial cavity permits identifica-        detecting intracavitary lesions, such
     tive predictive value and a 99.7%     tion of focal lesions, polyps, fibroids,    as polyps and fibroids, than is blind
     negative predictive value.            dyssynchronous endometrium, hyper-          endometrial biopsy alone. A recent
•    Fleischer et al27 used TVUS to        plasia or cancer. If a focal lesion is      study of 181 patients reported a sen-
     evaluate 1,750 women without          identified, then a targeted endome-         sitivity of 96.6% and a specificity of
     PMB for a selective estrogen-         trial biopsy (with full removal of the      100% when hysteroscopy was com-
     receptor modulator study. When        lesion) is the most sensitive method of     bined with endometrial biopsy.31 Hys-
     the endometrial echo was < 5 mm       determining the cause of the increased      teroscopists sometimes find it difficult
     the negative predictive value was     endometrial thickness. If a diffusely       to distinguish between a proliferative,
     99.94% for excluding malignancy       thickened endometrium is identified, a      exaggerated endometrium and en-
     (1 cancer in 1,750 women) and         blind aspiration biopsy (with a pipelle     dometrial hyperplasia.32 This is why it
     99.77% for complex hyperplasia        or Vabra aspirate) provides a sensitive     is critical to also perform endometrial
     (4 in 1,750 women).                   analysis. If endometrial biopsy yields      biopsy with hysteroscopy. Theoreti-
•    Gull et al28 found that only 7 can-   insufficient tissue that is inadequate      cally, the specificity and positive pre-
     cers in 1,361 women were found        for evaluation, or if no endometrial        dictive value of hysteroscopy in cases
     using a 4-mm cutoff.                  tissue is obtained or stenosis is en-       of abnormal uterine bleeding should
                                           countered, direct visualization of the      be 100%. In practice, however, the
   The 4mm cutoff. The data are very       endometrium should be performed             false-negative rate is 2%–4%, and is
solid for the use of TVUS in the eval-     and directed biopsies taken.                the result of operator error in detect-
uation of PMB. Literature review sup-         Hysteroscopy. Hysteroscopy, prefer-      ing abnormal endometrial lesions.
ports a conservative approach when         ably office-based, is excellent for eval-       Hysteroscopic evaluation of the en-
the endometrial echo is < 4mm in a         uating the endometrial cavity and           dometrium in postmenopausal women
postmenopausal woman; the risk of          endocervix; the false-negative rate is      must be regimented. The hystero-
missing an endometrial cancer is low.      3%29 and it is more accurate than           scopist should systematically evaluate
However, to rely upon TVUS the en-         TVUS in detecting focal disease and         the endometrial height, surface and
dometrial echo should be fully viewed,     also has greater specificity. A review of   vascular architecture. Uterine disten-


tion media can include CO2 or saline;    ally single and are usually benign.          the endometrium overlying them be-
however, both may be complementary       Polyps respond to estrogen and ta-           comes markedly atrophic. The best
and used concurrently. The endome-       moxifen with growth, whereas they            view of an intramural fibroid appears
trial thickness is usually <5 mm in      are less sensitive to progesterone. As       when visualized from the internal os.
the postmenopausal woman. The en-        polyps grow, they can develop a pedi-        Hysteroscopy is an excellent method-
dometrial thickness can be assessed by   cle and even protrude into the endo-         ology to determine their number, size
pressing the tip of the hysteroscope     cervix or into the vagina. Polyps share      and location, and to determine if the
into the endometrium. Without estro-     the same surface as the surrounding          patient has an intracavitary fibroid. A
gen, there should be no compression or   endometrium. Polyps may have a va-           pedunculated fibroid can be hystero-
indentation of the endometrium; with     riety of colors, ranging from that           scopically resected completely. The
estrogenic stimulation, there may be a   which resembles the endometrium to           risk of sarcomatous changes is less
wave-like surface with pseudopolypoid    red-yellow, and at the distal tip of the     than 1% in patients with uterine fi-
projections noted.                       polyp an ecchymotic purple-blue color        broids (Figure 6).
                                         may be seen (Figure 5).                         Endometrial hyperplasia. Making
Evaluation Findings:                        Submucosal f ibroids. Intracavitary       the diagnosis of endometrial hyper-
Clinical characteristics                 fibroids are firm and protrude from          plasia in the postmenopausal woman
Proper diagnosis depends upon the        the endometrium. They are usually            is less difficult than in reproductive-
ability to accurately interpret find-    solitary, but may be accompanied by          age patients, who have a variable ap-
ings from the diagnostic tools dis-      additional leiomyomas. In menopause,         pearance of the endometrium due to
cussed above.                            the endometrium covering the fibroid         hormonal fluctuations. Even so, the
   Atrophy. Endometrial atrophy is       is thin. Blood vessels can be seen           diagnosis of hyperplasia is the most
the most common cause of bleeding        coursing over the surface of the my-         often missed diagnosis with hys-
during menopause. Additionally, en-      oma. On occasion, ulceration of the          teroscopy, compared to SIS.33 The
dometrial atrophy is common in peri-     endometrium opposite the myoma               most likely reason for “missing”the hy-
menopausal women who use low-dose        can be seen. When the myoma is pal-          perplasia may be due to the techniques
hormonal contraception. DepoProvera,     pated with the distal tip of the hys-        employed. When fluid distention or
DepoLupron, danazol, progesterone-       teroscope, resistance is met and, unlike     CO2 compresses the endometrium, the
only intrauterine systems or proges-     a polyp, the myoma cannot be pushed          endometrial projections of hyperplastic
terone-only oral contraception can       away from the hysteroscope. Intra-           tissue are more difficult to discern.
also be associated with atrophy. At-     mural fibroids are unlikely sources of       Making the diagnosis of endometrial
rophic endometrium has a translucent     bleeding during menopause unless             hyperplasia requires close surveillance
and porcelain appearance. The sur-
face is smooth and dull. An atrophic
endometrium may have pale endo-
metrium covered by focal mucus-
containing cysts, which, in turn, are
covered by a thin endometrium. Dif-
fuse or isolated petechial hemorrhages
are frequently noted. Gland openings
or cystic atrophy with openings adja-
cent to the myometrium are also
noted. These gland openings appear
translucent blue-gray. Because the en-
dometrium is thin, the underlying
muscle bundles of the myometrium
and interlacing columns and recesses
may produce diverticula.                      Figure 5. In-office flexible hysteroscopic view utilizing CO2. Note the
   Endometrial polyps. Endometrial            endocervical myoma and an endometrial polyp arising from the left
polyps are a frequent cause of bleed-         lateral wall.
ing. Polyps may vary in size, are usu-

                                                                                              NOVEMBER /DECEMBER 2008      19

of the tissue thickness, color, vascula-
ture and consistency of the endo-
metrium. Hyperplasia can be focal or
global. Hyperplastic tissue has no
organized structure and is an out-
growth of aberrant tissue, so it is easily
friable and tears easily when touched
with the hysteroscope. Abnormal-
appearing endometrium may have fo-
cal lesions that include polyps. Addi-
tional cause for suspecting hyperplasia
includes focal or papillary mucosal
projections with or without gland
cysts, an abnormal vascular network
with atypical vessels, and crowded or              Figure 6. In-office flexible hysteroscopic view utilizing CO2, demonstrating
abnormally-spaced gland openings.34                a vascular submucosal fibroid.
   Endometrial cancer. Endometrial
cancer has myriad appearances. Espe-
cially among postmenopausal women,            proved imaging, Goldstein et al36            of 2 to 3 cases per 1,000 women, has
the diagnosis is less likely to be missed.    advocated SIS to monitor tamoxifen           been noted among tamoxifen users.37
Boring, pale, atrophic, monotonous            effects. With traditional TVUS, the
endometrial architecture best describes       endometrial thickness appears highly         Hormone Therapy and
the postmenopausal endometrium.               unusual and heterogeneous, and               Abnormal Bleeding
The chance of endometrial cancer in-          demonstrates centrally located uterine       Many women use HT for the allevi-
creases when the hysteroscopist detects       changes; if SIS is not performed, these      ation of vasomotor symptoms, to de-
thickened, irregular endometrium; in-         unusual features can easily be over-         crease the risk of osteoporosis and
creased surface vascularity; friable cells;   interpreted. Unlike TVUS, SIS more           improve urogenital atrophy. In the
or when there is difficulty distending        accurately detects endometrial health        Women’s Health Initiative study of
the endometrium.                              and can determine if additional im-          over 8,000 women randomized to HT,
   Endometrial cancer requires histo-         aging is necessary. SIS can identify         about 40% of women complained of
logic diagnosis; however, the index of        the subendometrial sonolucencies             abnormal bleeding on the combined,
suspicion can be raised when the to-          to the proximal myometrium. The              continuous HT regimen utilizing
pography of the endometrium is ir-            abnormalities may represent abnor-           conjugated equine estrogen 0.625 mg
regular or has focal lesions. Sugimoto        mal adenomyomatous-like changes              and medroxyprogesterone acetate 2.5
et al35 have noted a high sensitivity in      in the proximal myometrium that              mg/daily.38 Users of estrogen/proges-
detecting endometrial cancer when             are microcysts. When viewed micro-           terone therapy (EPT) have a lower
the following features are observed:          scopically, the junction between the         risk of endometrial cancer than do
papillary, polypoid, nodular or mixed         endometrium and myometrium is ir-            nonusers. Current recommendations
endometrial growth demonstrating              regular and nonlinear, whereas in pa-        include watchful waiting if vaginal
friable tissue, focal necrosis and atyp-      tients not receiving tamoxifen, the          bleeding occurs during the first 3
ical vessels.                                 junction is linear.                          months of EPT. As discussed previ-
   Tamoxifen-induced changes. Endo-              Other ultrasonographic findings           ously, in-office hysteroscopy or TVUS
metrial surveillance of asymptomatic          observed in patients taking tamoxifen        can be used to evaluate bleeding if it is
patients taking tamoxifen does not            (versus controls) consist of increased       persistent. Additionally, if patients are
differ from that of women routinely           uterine volume and depth, greater            unwilling to continue therapy because
using ET. Although most long-term             endometrial thickness, increased inci-       of bleeding, then hysteroscopy or SIS
users of tamoxifen will have an inac-         dence of endometrial polyps (36% vs          can be recommended. As already
tive endometrium, some will show in-          10%) and increased endometrial atro-         stated, if TVUS only is used and the
creased endometrial thickness during          phy (28% vs 87%). A slight increase          endometrial echo is < 5 mm, SIS,
conventional TVUS. In light of im-            in the incidence of endometrial cancer,      when available, should be performed


next to exclude intracavitary lesions     cian do in cases of profound cervical       to remove the stenotic cervical os.
when bleeding persists.                   stenosis?                                   Rarely will the clinician need to resort
   Patients on cyclic combined EPT           Traditionally, laminaria has been        to hysterectomy for marked cervical
should receive a minimum of 12–13         used for cervical dilation. But even        stenosis and PMB; however, in the
days of progestin therapy. Endome-        the smallest laminaria cannot be            presence of an abnormal ultrasound
trial biopsy should be considered in      placed with marked cervical stenosis.       including a non-visualized endo-
patients who bleed before day 11 of a     Oral or vaginal misoprostol is espe-        metrium, thickened endometrium or
cyclic progestin regimen. If secretory
or pseudodecidual changes are not
evident, then additional progestin        In some cases, HT may produce bleeding in women who
should be used. Patients using cyclical   were previously asymptomatic.
EPT who experience bleeding should
be reassured if their bleeding occurs
during the final 12–14 days of the        cially effective for cervical stenosis.     cervical dysplasia, hysterectomy is a
progestin regimen or during the week      Oral misoprostol (200-400 mcg 48            reasonable option for managing ab-
following progestin withdrawal. If        hours and, again, 8–12 hours before         normal bleeding when cervical steno-
erratic bleeding occurs, then evalua-     cervical dilation) is associated with       sis is encountered.43 In a retrospective
tion with in-office hysteroscopy or       increasing cervical softening and en-       study of patients with PMB who had
SIS should be undertaken.                 hancing placement of a dilator.41 Pa-       cervical stenosis that precluded fur-
   Whichever HT formulation is            tients may have transient cramps,           ther evaluation and were subsequently
chosen, approximately 30%–40% of          diarrhea or low-grade fever; however,       treated by hysterectomy, 64% had
women initiating HT will experience       misoprostol greatly facilitates per-        benign pathology, 12% had cervical
episodic or prolonged bleeding while      formance of the procedure.                  dysplasia (12%) and 4% had endome-
on therapy.39 In some cases, HT may           When misoprostol or laminaria           trial cancer.43
produce bleeding in women who were        attempts are not helpful, dilation of
previously asymptomatic. Earlier in-      the cervix under ultrasound guidance        Summary
tervention with in-office hysteroscopy    should be considered. Ultrasound            Understanding the normal variations
or SIS demonstrating endometrial at-      confirms proper placement of the in-        of perimenopausal bleeding is impor-
rophy or the presence of intracavitary    struments into the uterine cavity.          tant. Likewise, postmenopausal women
lesions may allay many fears. If endo-    Transabdominal ultrasound greatly           warrant thorough evaluation of the
metrial atrophy is detected, the pa-      improves the ability to guide an in-        endometrium when PMB occurs. For
tient can be reassured that bleeding      trauterine catheter or endometrial          both populations a low threshold for
will likely resolve with continued HT     pipelle biopsy device into the uterine      endometrial assessment is essential.
usage. Benign polyps and fibroids may     cavity, and diminishes the likelihood of    Clinical tools used to evaluate the
continue to bleed, requiring hystero-     creating a false tract or perforation or    endometrium include endometrial
scopic resection. Likewise, if a pre-     abandoning the procedure. If this can-      biopsy, in-office hysteroscopy, TVUS
malignant or occult malignancy is         not be performed in the office, it can      and SIS.The use of cervical-softening
detected, prompt treatment with           be performed under light sedation.42        agents, such as misoprostol, helps to
progesterone therapy or hysterectomy          Alternatively, flexible hysteroscopy    make hysteroscopy, SIS and endome-
should be undertaken.                     may be more advantageous when a             trial biopsies more comfortable, de-
                                          tortuous uterine cavity is encountered.     creases the risk of uterine perforation
Clinical Conundrums                       If a 3-mm flexible hysteroscope can be      or cervical lacerations, and facilitates
For the woman who presents with           placed into the cervix, it may be able to   cervical dilation. Hysteroscopy, par-
PMB and who has marked cervical           be navigated through the endocervix         ticularly with small-diameter hystero-
stenosis, evaluation can be challeng-     and into the uterine cavity to obtain       scopes, provides excellent imaging of
ing. Prior surgical procedures, such      visualization of the endometrium.The        the endocervix and endometrium in
as laser conizations, loop excision       use of fluid or CO2 also helps dilate the   the office setting. When hysteroscopy
and cold-knife biopsy, are associated     cervix. When marked cervical stenosis       is not available,TVUS is helpful in the
with a 1%–40% risk of cervical            is encountered consider a shallow-cone      initial triage of patients with abnormal
stenosis.40 But what can the clini-       loop electrosurgical excision procedure     uterine bleeding. TVUS, however,

                                                                                              NOVEMBER /DECEMBER 2008       21

                                                           and postmenopausal women with abnormal uterine                  25. Smith-Bindman R, Kerlikowske K, Feldstein VA, et al.
does have limitations. When the en-                        bleeding. Menopause 2007;14:846-52.                             Endovaginal ultrasound to exclude endometrial cancer
dometrium is not well visualized with                      7. Breitkopf DM, Frederickson RA, Snyder RR. Detec-             and other endometrial abnormalities. JAMA 1998;
                                                           tion of benign endometrial masses by endometrial stripe         280:1510-17.
TVUS or bleeding continues despite                         measurement in premenopausal women. Obstet Gynecol              26. Tsuda H, Kawabata M, Yamamoto K, et al. Prospec-
normal endometrial parameters, hys-                        2004;104:120-5.                                                 tive study to compare endometrial cytology and trans-
                                                           8. de Kroon CD, de Bock GH, Dieben SW, Jansen FW.               vaginal ultrasonography for identification of endometrial
teroscopy or SIS is advisable. The ar-                                                                                     malignancy. Gynecol Oncol 1997;65:383-6.
                                                           Saline contrast hysterosonography in abnormal uterine
ray of clinical tools available can direct                 bleeding: a systematic review and meta-analysis Br J            27. Fleischer A, Wheeler JE, Lindsay I, et al. An assess-
                                                           Obstet Gynaecol 2003;110:938-47.                                ment of the value of ultrasonographic screening for en-
therapy, and provide guidance for ini-                                                                                     dometrial disease in postmenopausal women without
                                                           9. Astrup K, Olivarius Nde F. Frequency of sponta-
tiation of surgery or expectant man-                       neously occurring postmenopausal bleeding in the gen-           symptoms. Am J. Obstet Gynecol 2001;184:740-4.

agement. ■                                                 eral population. Acta Obstet Gynecol Scand 2004;                28. Gull B, Carlsson SA, Karlsson B, et al. Transvaginal
                                                           83:203-7.                                                       ultrasonography of the endometrium in women with post-
                                                           10. Eitan R, Saenz CC, Venkatraman ES, et al. Pilot study       menopausal bleeding: is it always necessary to perform
                                                                                                                           an endometrial biopsy? Am J Obstet Gynecol 2000;
Linda D. Bradley, MD, is Vice Chairman                     prospectively evaluating the use of the measurement of
                                                           preoperative sonographic endometrial thickness in post-
of Obstetrics & Gynecology, Women’s                        menopausal patients with endometrial cancer. Menopause          29. Clark TJ, Voit D, Gupta JK, et al. Accuracy of hys-
Health Institute; and Director, Center                     2005;12:27-30.                                                  teroscopy in the diagnosis of endometrial cancer and
                                                                                                                           hyperplasia: a systematic quantitative review. JAMA
for Menstrual Disorders, Fibroids, and                     11. Marchetti M, Vasile C, Chiarelli S. Endometrial can-
                                                           cer: asymptomatic endometrial findings. Characteristics
Hysteroscopic Services, Cleveland                          of postmenopausal endometrial cancer. Eur J Gynae-              30. De Jong P, Doel F, Falconer A. Outpatient diagnostic
Clinic Department of Obstetrics and                        colog Oncol 2005;26:479-84.                                     hysteroscopy. Br J Obstet Gynaecol 1990;97;299-303.
Gynecology, Cleveland, OH.                                 12. Gredmark T, Kvint S, Havel G, et al. Histopathologi-        31. Marchetti M, Litta P, Lanza P, et al. The role of hys-
                                                           cal findings in women with postmenopausal bleeding.             teroscopy in early diagnosis of endometrial cancer. Eur
Malgorzata E. Skaznik-Wikiel, MD, is                       Br J Obstet Gynaecol 1995;102:133-6.                            J Gynaecol Oncol 2002;23:151–3.

Resident, Obstetrics and Gynecology,                       13. Karlsson B, Granberg S, Wikland M, et al. Trans-            32. Fay TN, Khanem N, Hosking D. Out-patient hys-
                                                           vaginal ultrasongraphy of the endometrium in women              teroscopy in asymptomatic postmenopausal women.
Case Western Reserve University,                           with postmenopausal bleeding—a Nordic multicenter               Climacteric 1999;2:263–7.
Cleveland Clinic/MetroHealth Medical                       study. Am J Obstet Gynecol 1995;172:1488-94.                    33. Widrich T, Bradley LD, Mitchinson A, Collins R.
Center, Cleveland Clinic Department                        14. Goldstein SR, Nachtigall M, Snyder JR, Nachtigall L.        Comparison of saline infusion sonography with office
                                                           Endometrial assessment by vaginal ultrasonography be-           hysteroscopy for the evaluation of the endometrium.
of Obstetrics and Gynecology, Cleve-                       fore endometrial sampling in patients with post-                Am J Obstet Gynecol 1996;174:1327-34.
land, OH.                                                  menopausal bleeding. Am J Obstet Gynecol 1990;                  34. Garuti G, Sambruni I, Colonnelli M, Luerti M. Accu-
                                                           163:119-23.                                                     racy of hysteroscopy in predicting histopathology of
   Dr. Bradley reports grant/research                      15. Gull B, Karlsson B, Milsom I, Granberg S. Can ultra-        endometrium in 1500 women. J Am Assoc Gynecol
                                                           sound replace dilation and curettage? A longitudinal            Laparosc 2001;8:207-13.
support from Microsulis and Conceptus.                     evaluation of postmenopausal bleeding and transvaginal          35. Sugimoto O. Hysteroscopic diagnosis of endometrial
She is a consultant for Biosphere, Ameri-                  sonographic measurement of the endometrium as predic-           carcinoma. A report of fifty-three cases examined at the
can Medical Systems, Gyrus ACMI and                        tors of endometrial cancer. Am J Obstet Gynecol 2003;           women’s clinic of Hyoto University Hospital. Am J Obstet
                                                           188:401-8.                                                      Gynecol 1975;121:105-13.
Ethicon Women's Health and Urology
                                                           16. Rose PG. Endometrial carcinoma. N Engl J Med                36. Goldstein SR, Nachtigall M, Snyder JR, Nachtigall L.
(Gynecare), and serves on the speakers'                    1996;335:640-9.                                                 Endometrial assessment by vaginal ultrasonography be-
bureaus for Wyeth and Novartis.                            17. Epstein E, Ramierez A, Skoog L, et al. Transvaginal         fore endometrial sampling in patients with postmeno-
   Dr. Skaznik-Wikiel reports no poten-                    sonography, saline contrast sonohysterography and               pausal bleeding. Am J Obstet Gynecol 1990;163:
tial conflicts related to the content of                   hysteroscopy for the investigation of women with post-          119-23.
                                                           menopausal bleeding and endometrium > 5 mm.                     37. Fisher B, Costantino JP, Wickerham DL, et al. Tamox-
this article.                                              Ultrasound Obstet Gynecol 2001;18:157-62.                       ifen for prevention of breast cancer—report of the Na-
   Submitted: September 21, 2007                           18. Twu NF, Chen SS. Five-year follow-up of patients            tional Surgical Adjuvant Breast and Bowel Project. P-1
Accepted: January 13, 2008                                 with recurrent postmenopausal bleeding. Zhonghua Yi             Study. J Natl Cancer Inst 1998;90:1371-88.
                                                           Xue ZaZhi (Taipei) 2000;63:628-33.                              38. Writing Group for the Women’s Health Initiative In-
References                                                 19 Chhieng DC, Elert P, Cohen JM, Cagiarella JF. Clini-         vestigators. Risks and benefits of estrogen plus progestin
1. Treloar AE. Menstrual cyclicity and the pre-            cal significance of atypical glandular cells of undeter-        in healthy postmenopausal women: principal results from
menopause. Maturitas 1981;3:249-64.                        mined significance in postmenopausal women. Cancer              the Women’s Health Initiative randomized controlled
                                                           2001;93:1-7.                                                    trial. JAMA 2002;288:321-33.
2. Seltzer VL, Benjamin R, Deutsch S. Perimenopausal
bleeding patterns and pathologic findings. J Am Med        20. Epstein E, Ramirez A, Skoog L, et al. Dilation and          39. Ettinger B, Pressman A, Bradley C. Comparison of
Womens Assoc 1990;45:132-4.                                curettage fails to detect most focal lesions in the uterine     continuation of postmenopausal hormone replacement
                                                           cavity in women with postmenopausal bleeding. Acta              therapy: transdermal vs oral estrogen. Menopause
3. Landgren BM, Collins A, Csemiczky G, et al.
                                                           Obstet Gynecol Scand 2001;80:1131-36.                           1998;5:152-6.
Menopause transition: annual changes in serum hor-
monal patterns over the menstrual cycle in women during    21. Moodley M, Roberts C. Clinical pathway for the              40. Baldauf JJ, Dreyfus M, Ritter J, et al. Risk of cervical
a nine-year period prior to menopause. J Clin Endocrinol   evaluation of postmenopausal bleeding with an empha-            stenosis after large loop excision or laser conization.
Metab 2004;89:2763-69.                                     sis on endometrial cancer detection. J Obstet Gynaecol          Obstet Gynecol 1996;88:933-8.
                                                           2004;24:736-41.                                                 41. Thomas JA, Leyland N, Durand N, et al. The use of
4. Farquhar C, Ekeroma A, Furness S, Arroll B. A sys-
tematic review of transvaginal ultraonography, sonohys-    22. Clark TJ, Barton PM, Coomarasamy A, et al. Investi-         oral misoprostol as a cervical ripening agent in operative
terography and hysteroscopy for the investigation of       gating postmenopausal bleeding for endometrial cancer:          hysteroscopy: a double-blind, placebo-controlled trial.
abnormal uterine bleeding in premenopausal women.          cost-effectiveness of initial diagnostic strategies. Br J Ob-   Am J Obstet Gynecol 2002;186:876-9.
Acta Obstet Gynecol Scand 2003;82:493-504.                 stet Gynaecol 2006;113:502-10.                                  42. Weiderpass E, Adami HO, Baron JA, et al. Risk of
5. Schwarzler P, Concin H, Bosch H, et al. An evalua-      23. Shushan A, Revel A, Rojansky N. How often are               endometrial cancer following estrogen replacement with
tion of sonohysterography and diagnostic hysteroscopy      endometrial polyps malignant? Gynecol Obstet Invest             and without progestins. J Natl Cancer Inst 1999;1131-37.
for the assessment of intrauterine pathology. Ultrasound   2004;58:212-15.                                                 43. Newman C, Finan M. Hysterectomy in women with
Obstet Gynecol 1998;11:337-42.                             24. Savelli L, De Iaco P, Santini D, et al. Histopathologic     cervical stenosis. J Reprod Med 2003;48:672-6.
6. Erdem M, Bilgin U, Bozhurt N, et al. Comparison of      features and risk factors for benignity, hyperplasia, and
transvaginal ultrasonograpy and saline infusion sonohys-   cancer in endometrial polyps. Am J Obstet Gynecol
terography in evaluating the endometrial cavity in pre-    2003;188:927-31.


Shared By: