Evaluation and Management of Adnexal Masses.pdf

The American College of Obstetricians and Gynecologists WOMEN’S HEALTH CARE PHYSICIANS P RACTICE BULLET IN clinical management guidelines for obstetrician – gynecologists Number 174, November 2016...

The American College of Obstetricians and Gynecologists WOMEN’S HEALTH CARE PHYSICIANS P RACTICE BULLET IN clinical management guidelines for obstetrician – gynecologists Number 174, November 2016 (Replaces Practice Bulletin Number 83, July 2007) Evaluation and Management of Adnexal Masses Adnexal masses (ie, masses of the ovary, fallopian tube, or surrounding tissues) commonly are encountered by obstetrician–gynecologists and often present diagnostic and management dilemmas. Most adnexal masses are detect- ed incidentally on physical examination or at the time of pelvic imaging. Less commonly, a mass may present with symptoms of acute or intermittent pain. Management decisions often are influenced by the age and family history of the patient. Although most adnexal masses are benign, the main goal of the diagnostic evaluation is to exclude malig- nancy. The purpose of this document is to provide guidelines for the evaluation and management of adnexal masses in adolescents, pregnant women, and nonpregnant women and to outline criteria for the identification of adnexal masses that are likely to be malignant and may warrant referral to or consultation with a gynecologic oncologist. Background sharply after the onset of menopause (1). According to data reported by the Surveillance, Epidemiology, and End Results Differential Diagnosis program, from 2009 to 2013, the median age at ovarian A pelvic mass can have gynecologic or nongynecologic cancer diagnosis was 63 years, and 69.4% of patients were origins (Box 1). Consideration of the location of a pelvic 55 years or older (1). Most adnexal masses in postmenopaus- mass in conjunction with patient age and reproductive al women are benign neoplasms, such as cystadenomas, but status can help narrow the differential diagnosis. Adnexal the risk of malignancy is much greater than in premenopausal masses of gynecologic origin may be benign or malig- women (2). nant ovarian lesions; tubal or paratubal processes such as The most important personal risk factor for ovarian hydrosalpinges or ectopic pregnancy; and uterine abnor- cancer is a strong family history of breast or ovarian can- malities such as leiomyomas or müllerian abnormalities. cer (3). It is important to distinguish a family history of Nongynecologic causes of pelvic masses are less common ovarian cancer from a familial ovarian cancer syndrome. and may be related to a variety of other organ systems, For a 35-year-old woman with one affected family mem- including gastrointestinal and urologic sources. Cases of ber, the lifetime probability of ovarian cancer increases metastatic cancer, especially those from the breast, colon, from a general population risk of 1.6% to a risk of 5% or stomach, may first present as adnexal masses. (4). However, for a woman with a BRCA1 mutation, the lifetime risk of ovarian cancer, fallopian tube cancer, Risk Factors for Malignancy or peritoneal cancer is approximately 41–46% by age Age is the most important independent risk factor for ovarian 70 years (5–8). For a woman with a BRCA2 mutation, cancer in the general population, with the incidence increasing the lifetime risk of ovarian cancer, fallopian tube cancer, Committee on Practice Bulletins—Gynecology. This Practice Bulletin was developed by the American College of Obstetricians and Gynecologists’ Committee on Practice Bulletins—Gynecology in collaboration with Ramez Eskander, MD; Michael Berman, MD; and Lisa Keder, MD, MPH. The information is designed to aid practitioners in making decisions about appropriate obstetric and gynecologic care. These guidelines should not be construed as dictating an exclusive course of treatment or procedure. Variations in practice may be warranted based on the needs of the individual patient, resources, and limitations unique to the institution or type of practice.  uncertain, and likely malignant, which can then guide appropriate management. Box 1. Differential Diagnosis of an Adnexal Mass ^ Medical and Family History Gynecologic A personal medical history with a detailed gynecologic Benign history and review of symptoms are critical compo- − Functional cyst nents of patient evaluation. In addition, a family history − Endometrioma and a review of other risk factors will help assess the likelihood of malignancy. See Committee Opinion No. − Tubo-ovarian abscess 478, Family History as a Risk Assessment Tool, and − Mature teratomas (dermoids) the National Comprehensive Cancer Center’s guide- − Serous cystadenoma lines, Genetic/Familial High-Risk Assessment: Breast − Mucinous cystadenoma and Ovarian, for details on how to take a relevant and − Hydrosalpinx detailed family history (13, 14). − Paratubal cysts Patients with adnexal masses may present with symptoms that can refine the differential diagnosis. − Leiomyomas For example, a woman of reproductive age with acute − Müllerian anomalies onset abdominal or pelvic pain may have a hemorrhagic Malignant or bleeding ovarian cyst. The potential for pregnancy − Epithelial carcinoma should be evaluated in all women of reproductive age − Germ cell tumor because ectopic pregnancy is in the differential diagnosis − Metastatic cancer of an adnexal mass in early pregnancy. Symptoms of uni- lateral, intermittent, and then acutely worsening pelvic − Sex-cord or stromal tumor pain may indicate an ovarian torsion. A more indolent, Nongynecologic progressive pelvic pain associated with fevers, chills, Benign vomiting, and vaginal discharge may indicate an infec- tious etiology such as a tubo-ovarian abscess. Women − Diverticular abscess who report acute or chronic dysmenorrhea or pain with − Appendiceal abscess or mucocele intercourse may have an endometrioma. Persistent bloat- − Nerve sheath tumors ing, generalized abdominal pain, and early satiety may − Ureteral diverticulum be signs of malignancy (15). Abnormal uterine bleeding − Pelvic kidney or postmenopausal bleeding may be caused by estrogen − Bladder diverticulum produced by sex cord-stromal tumors (16). Malignant Physical Examination − Gastrointestinal cancers The physical examination should start with evalua- − Retroperitoneal sarcomas tion of vital signs and general physical appearance. − Metastatic cancer Whether the woman has a symptomatic adnexal mass or one that is incidentally discovered on imaging, a comprehensive physical examination should include palpation of cervical, supraclavicular, axillary and or peritoneal cancer is 10–27% by age 70 years (5–8). groin lymph nodes; pulmonary auscultation; abdomi- The risk of ovarian cancer through age 70 years for nal palpation and auscultation; and pelvic examination women with Lynch syndrome is estimated to be 5–10% (including visual inspection of the perineum, cervix, (7–9). Additional factors that increase ovarian cancer and vagina; and bimanual palpation, with rectovaginal risk include nulliparity, early menarche, late menopause, examination as indicated). Although pelvic examina- white race, primary infertility, and endometriosis (10–12). tion (even with the patient under general anesthesia) has shown limited ability to identify an adnexal mass, General Evaluation especially with patient body mass index greater than Individual patient characteristics, physical examination 30 (17), examination findings that are concerning for findings, imaging results, and serum marker levels help adnexal malignancy include a mass that is irregular, separate masses into the categories of probably benign, firm, fixed, nodular, bilateral, or associated with ascites. 2 Practice Bulletin No. 174 Benign conditions that can produce these findings which is a protein associated with epithelial ovarian include endometriosis, chronic pelvic infections, hemor- malignancies, but also frequently expressed at lower lev- rhagic corpus luteum, tubo-ovarian abscess, and uterine els by nonmalignant tissue. Elevation of CA 125 levels leiomyomas (18). may occur in endometriosis, pregnancy, pelvic inflam- matory disease, and in nongynecologic cancer. In evalu- Imaging ating adnexal masses, CA 125 measurement is most Transvaginal ultrasonography is the most commonly useful in postmenopausal women and in identifying used imaging technique for the evaluation of adnexal nonmucinous epithelial cancer (Table 1) (25). The CA masses. The ultrasound examination should assess the 125 level is elevated in 80% of patients with epithe- size and composition of the mass (cystic, solid, or lial ovarian cancer but in only 50% of patients with mixed); laterality; and the presence or absence of sep- stage I disease (25). More recently, human epididymis tations, mural nodules, papillary excrescences, or free protein 4 has been identified as a potentially useful bio- fluid in the pelvis. Spectral, color Doppler ultrasonogra- marker in distinguishing benign from malignant masses phy is useful to evaluate the vascular characteristics of (26, 27). If a less common ovarian histopathology is pelvic lesions (19). Abdominal ultrasonography is a use- suspected based on risk factors, symptoms, or ultrasound ful addition when pelvic structures are distorted by pre- findings, measurement of levels of ß-hCG, L-lactate vious surgery, when masses extend beyond the pelvis, dehydrogenase, alpha-fetoprotein, or inhibin may assist or if transvaginal ultrasonography cannot be performed. in the evaluation (Table 2). Computed tomography (CT), magnetic resonance Panels of biomarkers have been investigated to imaging (MRI), and positron emission tomography determine their ability to distinguish between benign (PET) are not recommended in the initial evaluation of and malignant adnexal masses when used in conjunction adnexal masses. Based on limited data, MRI may have superior ability compared with transvaginal ultraso- Table 1. Serum Biomarker and Multimodal Test Results nography in correctly classifying malignant masses at Considered Abnormal in Women With Adnexal Masses* ^ the expense of a lower overall detection rate (20, 21). However, MRI often is helpful in differentiating the Test Premenopausal Postmenopausal origin of pelvic masses that are not clearly of ovarian CA 125 –– † > 35 U/mL origin, especially leiomyomas (22, 23). MIA ≥ 5.0 ≥ 4.4 Currently, the best use of CT imaging is not to detect and characterize pelvic masses, but to evaluate the ROMA ≥ 1.31 ≥ 2.77 abdomen for metastasis when cancer is suspected based RMI > 200 > 200 on ultrasound images, examination results, or serum Abbreviations: CA, cancer antigen; MIA, multivariate index assay; ROMA, Risk of markers. A CT scan can detect ascites, omental metasta- Ovarian Malignancy Algorithm; RMI, risk of malignancy index. *Serum biomarker and multimodal testing may be helpful to identify a woman ses, peritoneal implants, pelvic or periaortic lymph node with an adnexal mass who would benefit from referral to or consultation with enlargement, hepatic metastases, obstructive uropathy, a gynecologic oncologist. Current evidence is insufficient to recommend any and possibly an alternate primary cancer site, including specific test. † pancreas or colon (24). Specificity and positive predictive value of CA 125 levels are consistently higher in postmenopausal women compared with premenopausal women. Prior American College of Obstetricians and Gynecologists’ guidance used a CA 125 Laboratory Testing threshold of greater than 200 U/mL for referral of a premenopausal woman with Laboratory testing may clarify the suspected etiology an adnexal mass to a gynecologic oncologist. This threshold was based on expert opinion; no evidence-based threshold is currently available; thus, gynecologic of a pelvic mass. Pregnancy testing should be obtained care providers should integrate the CA 125 level with other clinical factors in in reproductive-aged women, if indicated. If an infec- judging the need for consultation. tious etiology is suspected, a complete blood count and Data from Jacobs I, Bast RC Jr. The CA 125 tumour-associated antigen: a review testing for gonorrhea and chlamydial infection should of the literature. Hum Reprod 1989 Jan;4:1–12; Skates SJ, Mai P, Horick NK, Piedmonte M, Drescher CW, Isaacs C, et al. Large prospective study of ovarian be performed. Other laboratory tests that may have cancer screening in high-risk women: CA 125 cut-point defined by menopausal value depending on the history and examination findings status. Cancer Prev Res (Phila) 2011;4:1401–8; Bristow RE, Hodeib M, Smith A, include urinalysis, fecal blood testing or other assessment Chan DW, Zhang Z, Fung ET, et al. Impact of a multivariate index assay on refer- ral patterns for surgical management of an adnexal mass. Am J Obstet Gynecol of intestinal involvement, and serum marker testing. 2013; 209:581.e1-8; Jacobs I, Oram D, Fairbanks J, Turner J, Frost C, Grudzinskas JG. A risk of malignancy index incorporating CA 125, ultrasound and menopausal Serum Marker Testing status for the accurate preoperative diagnosis of ovarian cancer. Br J Obstet Serum markers are used in conjunction with imaging to Gynaecol 1990;97:922–9; and Moore RG, McMeekin DS, Brown AK, DiSilvestro P, Miller MC, Allard WJ, et al. A novel multiple marker bioassay utilizing HE4 assess the likelihood of malignancy. The most extensively and CA 125 for the prediction of ovarian cancer in patients with a pelvic mass. studied serum marker is cancer antigen 125 (CA 125), Gynecol Oncol 2009;112:40–6. Practice Bul­le­tin No. 174 3 with clinical and radiologic evaluation. The U.S. Food and Drug Administration has approved two different Clinical Considerations serum tumor marker panel tests to further assess the risk and Recommendations of ovarian cancer in adult women with pelvic masses: 1) the multivariate index assay, a qualitative serum tumor What is the role of ultrasonography in marker panel and 2) the Risk of Ovarian Malignancy the initial evaluation of a patient with a Algorithm (28, 29). These panels are approved for use suspected or an incidentally identified in women older than 18 years with an already identified adnexal mass? adnexal mass that requires surgery. The multivariate index assay incorporates five serum biomarkers that Transvaginal ultrasonography is the recommended imag- ing modality for a suspected or an incidentally identified are associated with ovarian cancer (CA 125 II, transfer- pelvic mass. No alternative imaging modality has demon- rin, transthyretin [prealbumin], apolipoprotein A-1, and strated sufficient superiority to transvaginal ultrasonogra- β 2-microglobulin) into a malignancy risk score of 0–10 phy to justify its routine use. using a proprietary algorithm (30). The Risk of Ovarian High frequency, gray-scale transvaginal ultrasonog- Malignancy Algorithm includes CA 125, human epi- raphy can produce high-resolution images of an adnexal didymis protein 4, and menopausal status (31). Variable mass that approximate its gross anatomic appearance. cut-off values specific to menopausal status have been Although image quality is operator dependent, interob- established for the multivariate index assay and the Risk server agreement among experienced ultrasonographers of Ovarian Malignancy Algorithm (Table 1). is high (35–37). The advantages of transvaginal ultra- Multimodal Tests sonography, including its widespread availability, good patient tolerability, and cost-effectiveness, make it the Diagnostic algorithms have been developed that incorpo- most widely used imaging modality to evaluate adnexal rate serum markers, clinical information, and ultrasound masses (38, 39). The main limitation of transvaginal findings. The United Kingdom’s National Institute ultrasonography alone as a diagnostic tool to distinguish for Health and Clinical Excellence Guidelines for the benign from malignant masses relates to its lack of recognition and initial management of ovarian cancer specificity and low positive predictive value for cancer, recommend the calculation of the risk of malignancy especially in premenopausal women (38–40). Color index version I as part of the evaluation (32). The risk Doppler ultrasonography permits measurement of blood of malignancy index is calculated using the product of flow in and around a mass and can increase the speci- the serum CA 125 level (U/mL), the ultrasound scan ficity of gray-scale two-dimensional ultrasonography result (expressed as a score of 0, 1, or 3 depending on (41, 42). However, the ranges of values of resistive index, findings), and the menopausal status (1 if premenopausal pulsatility index, and maximum systolic velocity between and 3 if postmenopausal) (33). A systematic review benign and malignant masses overlap considerably (43). found a risk of malignancy index score of 200 (Table In an attempt to overcome the overlap, three-dimensional 1) to have a pooled estimated sensitivity of 78% (95% ultrasound examination of vascular architecture has been confidence interval [CI], 71–85%) and a specificity of used and proved discriminatory in distinguishing benign 87% (95% CI, 83–91%) (34). masses from cancer in some reports (44, 45). What ultrasound findings suggest malignancy? Table 2. Serum Biomarkers in Ovarian Germ Cell Tumors ^ Ultrasound findings that should raise the clinician’s β-hCG AFP LDH CA 125 level of concern regarding malignancy include cyst size greater than 10 cm, papillary or solid components, Dysgerminoma + - + - irregularity, presence of ascites, and high color Doppler Endodermal sinus - + - - flow. There has been significant research on the use of tumo ultrasound scoring systems alone or in combination with Choriocarcinoma + - - - serum markers or historical information for predicting Immature teratoma - + + + malignancy. Although promising, these systems have Embryonal carcinoma + + - - been validated only in research settings with specific Abbreviations: AFP, alpha fetoprotein; CA, cancer antigen; LDH, lactate ultrasound training and their suitability for routine clini- dehydrogenase. cal use has not been fully clarified. 4 Practice Bulletin No. 174  The International Ovarian Tumor Analysis (IOTA) Simple cysts are almost always universally benign, group has incorporated ultrasound features into its regardless of menopausal status or cyst size, with malig- Logistic Regression model 2 and its Simple Rules, which nancy rates in most series of 0–1% (41, 51–53, 57, were designed to help ultrasonographers predict the risk 58). In the largest prospective study published to date, of malignancy of an adnexal mass before surgery (46, 2,763 postmenopausal women with unilocular cysts 47). The IOTA Logistic Regression model 2 includes six no larger than 10 cm were evaluated using serial variables (patient age and five ultrasound findings sug- transvaginal ultrasonography at 6-month intervals (53). gestive of malignancy) that are entered into a formula that Spontaneous resolution occurred in more than two thirds calculates the probability of malignancy (46). The IOTA of patients, and no cases of cancer were detected after a Simple Rules include a total of 10 ultrasound findings mean follow-up of 6.3 years. A more recent series exam- characteristic of malignant adnexal masses and benign ined the risk of malignancy in 1,148 masses classified adnexal masses, with accompanying guidance on how to as unilocular cysts on ultrasonography (59). Of these apply these rules (47). cases, 11 (0.96% [95% CI, 0.48–1.71]) were malignant; Several other transvaginal ultrasound scoring sys- however, in seven of the 11 malignancies, the ultrasound tems have been developed that quantify cancer risk assessment (59) did not detect papillary projections or based on morphology (48). Whereas scoring criteria other solid components that subsequently were found vary among the various models, most assign low risk macroscopically at surgery. scores to sonolucent cysts with smooth walls, thin or Small descriptive studies have reported ultrasound absent septations, and absence of solid components. characteristics that may be specific for selected benign In general, the various morphologic ultrasound scor- diagnoses. Typical findings reported for endometriomas ing systems are able to distinguish benign from malig- include a round homogeneous-appearing cyst contain- nant masses in most instances (48). However, a 2014 ing low-level echoes within the ovary (60). These systematic review and meta-analysis that compared characteristics allow differentiation from other types of various malignancy prediction models (including ultra- ovarian cysts with sensitivity of 83% and a specificity sound morphologic scoring systems, the IOTA Logistic of 89% and a positive and negative predictive value of 77% and 92%, respectively (61). Mature teratomas, Regression model 2, the IOTA Simple Rules, biomarker which contain a hypoechoic attenuating component with panels, and various versions of the risk of malignancy multiple small homogeneous interfaces, were deter- index multimodal test) found that the best performing mined with 98% accuracy in a series of 155 suspected models were the IOTA Logistic Regression model 2 dermoid cysts (62). Overall, ultrasound assessment has with a risk cut-off of 10% and the IOTA Simple Rules shown a reported sensitivity of 58% and specificity of model (49). The IOTA Logistic Regression model 2 and 99% in the diagnosis of mature cystic teratomas (63). the IOTA Simple Rules model demonstrated high sensi- Hydrosalpinges are another benign adnexal mass, which tivity (0.92 [95% CI, 0.88–0.95] for Logistic Regression on transvaginal ultrasonography appear as tubular-shaped model 2; 0.93 [95% CI, 0.89–0.95] for Simple Rules) sonolucent cysts, with a sensitivity of 93% and specific- and specificity (0.83 [95% CI, 0.77–0.88] for Logistic ity of 99% for differentiating them from other adnexal Regression model 2; 0.81 [95% CI, 0.76–0.85] for masses (64). Simple Rules) based on pooled data (49). What is the role of serum marker testing in What ultrasound findings suggest benign the initial evaluation of an adnexal mass? disease? Serum marker testing is indicated to evaluate the like- Ultrasound characteristics of benign masses include sim- lihood of malignancy and need for surgery. Elevated ple appearance: thin, smooth walls; and the absence of CA 125 levels in combination with other findings can solid components, septations, or internal blood flow on be useful to distinguish between benign and malignant color Doppler ultrasound imaging. These simple cysts adnexal masses and to identify patients who should be are highly likely to be benign in any age group (50–54). referred to or treated in consultation with a gynecologic A definitive size cutoff to delineate the need for surgi- oncologist. Specificity and positive predictive value cal intervention has not been established (52). Cysts of of CA 125 levels are consistently higher in postmeno- 10 cm or larger often are considered to be an indica- pausal women compared with premenopausal women tion for surgery (55); however, simple cysts (even those (65, 66). The combination of an elevated CA 125 level greater than 10 cm) often will spontaneously regress and a pelvic mass in a postmenopausal woman is highly when examined with serial ultrasonography (56). suspicious for malignancy, and patients with these Practice Bul­le­tin No. 174 5 findings should be referred to or treated in consultation suspected in women with a solid pelvic mass and irregu- with a gynecologic oncologist. Although CA 125 level lar or postmenopausal bleeding. measurement is less valuable in predicting cancer risk in premenopausal women than in postmenopausal women, What is the role of serum biomarker panel extreme values increase suspicion for a malignant pro- testing in the evaluation of an adnexal mass? cess. For example, although premenopausal women with adnexal masses and either normal or mildly elevated Serum biomarker panels may be used as an alternative to CA 125 levels usually have benign diagnoses, a mark- CA 125 level alone in determining the need for referral edly elevated CA 125 level raises greater concern for to or consultation with a gynecologic oncologist when an malignancy, even though women with benign condi- adnexal mass requires surgery. These biomarker panels tions such as endometriomas can have CA 125 level are not recommended for use in the initial evaluation elevations of 1,000 units/mL or greater (67). Prior of an adnexal mass, but may be helpful in assessing guidance of the American College of Obstetricians and which women would benefit from referral to a gyneco- Gynecologists used a CA 125 threshold of greater than logic oncologist. Trials that have evaluated the predic- 200 U/mL for referral of premenopausal women to gyne- tive value of these panels show potential for improved cologic oncologists. This threshold was based on expert specificity, especially for evaluation of premenopausal opinion. No evidence-based threshold is currently avail- women. However, comparative research has not yet able; thus, gynecologic care providers should integrate defined the best testing approach. the CA 125 level with other clinical factors in judging The multivariate index assay has demonstrated higher sensitivity and negative predictive value for ovar- the need for consultation. ian malignancy when compared with clinical impression The overall sensitivity of CA 125 testing in dis- and CA 125 alone (30, 74). In a study of 494 women tinguishing benign from malignant adnexal masses enrolled by nongynecologic oncology providers, the reportedly ranges from 61% to 90%; specificity ranges multivariate index assay correctly predicted ovarian from 71% to 93%, positive predictive value ranges malignancy in 91.4% (95% CI, 77.6–97.0) of cases of from 35% to 91%, and negative predictive value ranges early stage disease, compared with 65.7% (95% CI, from 67% to 90% (65, 68–72). Wide variations in these 49.2–79.2) for CA 125 alone (30). The multivariate figures reflect differences in cancer prevalence in the index assay was abnormal in 83.3% of malignancies in study population, the proportion of patients who are which the clinical impression was thought to be benign postmenopausal, and the threshold of CA 125 levels and was abnormal in 70.8% of cases of cancer in which considered abnormal. Cancer antigen 125 testing has the CA 125 was normal (30). In a larger cohort of a low sensitivity for the detection of ovarian cancer 1,016 patients, the multivariate index assay combined because the CA 125 level is elevated in only one half of with clinical assessment had greater sensitivity (95.3%; cases of early stage epithelial ovarian cancer and rarely 95% CI, 88.6–98.2) compared with clinical assessment in cases of germ cell, stromal, or mucinous cancer. Low or CA 125 alone for early-stage ovarian malignancies specificity occurs because the CA 125 level is elevated (74). The addition of radiologic imaging to the multi- in many nonmalignant clinical conditions, including variate index assay further increases sensitivity (98% for uterine leiomyomas, endometriosis, pelvic inflammatory ultrasonography and 97% for CT scan) and the negative disease, ascites of any etiology, and even inflamma- predictive value (99% for ultrasonography and 94% for tory conditions such as systemic lupus erythematosus CT scan). The false negative rate is less than 2% when and inflammatory bowel disease (73). Because most the results of imaging and the multivariate index assay of these clinical conditions occur in premenopausal indicate low risk (75). women and because most cases of epithelial ovarian The Risk of Malignancy Algorithm includes human cancer occur in postmenopausal women, the sensitiv- epididymis protein 4, which has been found to be more ity and specificity of an elevated CA 125 for cancer sensitive and specific than CA 125 for the evaluation diagnosis in the setting of a pelvic mass is highest after of adnexal masses (76). In a cohort of 531 patients, the menopause. Risk of Ovarian Malignancy Algorithm successfully Additional tumor marker testing may be useful classified patients into high-risk and low-risk groups, if a less common ovarian histopathology is suspected with 93.8% of cases of epithelial ovarian cancer classi- (Table 2). Levels of β-hCG, L-lactate dehydrogenase, fied as high risk before surgical exploration (31). In post- and alpha-fetoprotein may be elevated in the presence menopausal women, the Risk of Ovarian Malignancy of certain malignant germ cell tumors. Granulosa cell Algorithm had a specificity of 75% (95% CI, 66.9–81.4) tumors produce estrogen and inhibin and should be and a sensitivity of 92.3% (95% CI, 85.9–96.4) in 6 Practice Bulletin No. 174 distinguishing malignant pelvic masses (31). Conversely, What type of surgical intervention is appro- in premenopausal women the Risk of Ovarian Malignancy priate for a presumed benign adnexal mass? Algorithm score exhibited a specificity of 74.8% (95% CI, 68.2–80.6) and a sensitivity of 76.5% (95% CI, Minimally invasive procedures are the preferred route of 58.8–89.3) (31). In a prospective analysis that compared surgery for presumed benign adnexal masses. Regardless the efficacy of the multivariate index assay and the of the approach employed, fertility preservation should Risk of Ovarian Malignancy Algorithm in 146 patients be a priority when managing masses in adolescents and with surgically confirmed malignancies, the multivari- premenopausal women who have not completed child- ate index assay was found to be more sensitive (97% bearing. Even in women who present with large ovarian and 87%, respectively; P=.25). However, the Risk of cysts of 10 cm or greater, it is possible to save normal Malignancy Ovarian Algorithm was more specific than portions of the ovary and remove the cyst laparoscopi- the multivariate index assay (83% versus 55%, respec- cally (80–82). tively; P