January 2006
ANATOMIC PATHOLOGY
Ovarian Tumors
[Editorial]
Ovarian tumors, though relatively uncommon, are a frequent cause of alarm in both patients and clinicians, owing to the deadly nature of ovarian cancer and its ability to escape early detection. However, in the universe of ovarian tumors, epithelial malignancies are relatively rare, comprising only 20% of all neoplasms. Patient age correlates directly with malignant potential of ovarian tumors: two-thirds of benign tumors are diagnosed in women younger than 40 years of age, while 90% of ovarian cancers are observed after the age of 40. Overall, two-thirds of ovarian tumors are diagnosed during reproductive age. The universe of ovarian tumors and tumor-like lesions is a subject of entire books, so we have chosen to limit this issue of Pathology Case Reviews to recent developments in diagnosis and treatment, as well as especially difficult diagnostic problems of ovarian neoplasms.
Thorough understanding of clinical implications of pathologic diagnosis is crucial in ovarian neoplasia. Dr. Sandra Brooks, in her thorough review, discusses the issues clinicians face in the diagnosis and management of various types of ovarian tumors and the decision-making process, which starts with the pathology report.
The concept and terminology of borderline epithelial tumors of the ovary have been controversial for over a century, in spite of the acceptance of a borderline category in almost all current classifications of ovarian tumors. One of the most important recent breakthroughs in achieving a relative uniformity of our approach to borderline tumors was the Borderline Ovarian Tumor Workshop hosted by the NCI. Dr. Steven Silverberg, one of the directors of the workshop, presents a succinct and clear summary of the most controversial subjects addressed at the Workshop.
One of the most challenging areas in ovarian tumor pathology is the involvement of the ovary(ies) by metastatic tumor; these metastases are notorious mimickers of primary ovarian tumors. The most diagnostically problematic secondary ovarian tumors are those that originate in the large intestine, appendix, and pancreas. Metastases from these sites typically produce histologic patterns resembling primary ovarian endometrioid carcinoma or mucinous epithelial neoplasms of borderline and malignant types. Drs. R. Vang and B. Ronnett present a case of such a metastasis followed by a most thorough and lucid review of diagnostic pitfalls and features helpful in the diagnosis of metastases to the ovary. Recommendations on useful or potentially applicable immunohistochemical stains are also detailed.
In a most comprehensive review, Dr. G. McCluggage covers the most relevant and helpful immunohistochemical stains that aid in the differential diagnosis of ovarian neoplasms. Both common and rare tumors and their staining patterns are discussed, and recommendations are given on the useful panels of immunostains in select diagnostic categories.
Endometriosis, and especially atypical endometriosis, has been established as a risk factor for the development of carcinoma; the ovary is the most common site of this transformation. Although this pathogenetic relationship is well known, the diagnosis of atypical endometriosis remains a difficult and poorly reproducible feat; Dr. M. Fukunaga elucidates this diagnostic problem and reviews relevant literature.
An unusual ovarian tumor, and one of the few highly malignant ovarian neoplasms affecting young women, is the small cell carcinoma of the ovary, hypercalcemic type. Dr. Esther Oliva presents a case of this rare entity and discusses its morphologic diagnosis, as well as helpful ancillary techniques.
Another malignancy afflicting still younger patients is the ovarian yolk sac carcinoma. This tumor is extremely treatable; hence, it is of utmost importance to diagnose it correctly; this diagnosis can be challenging, however, since yolk sac tumor can have one or more of many morphologic patterns of which the pathologist must be aware in order to classify the lesion correctly. Drs. Rabban and Zaloudek describe all the morphologic patterns and diagnostic features of yolk sac tumors and their differential diagnosis.
Ovarian stromal tumors can present a diagnostic challenge when they extensively infarct or demonstrate increased mitotic activity and cellularity. Ioffe presents a case of a cellular (atypical) fibroma and discusses diagnostic criteria to distinguish it from a banal fibroma and the rare ovarian fibrosarcoma. Other tumors in this category are also covered.
Pathology Case Reviews, Volume 11(1), January/February 2006, pp 1-2
Ovarian Cancer: A Clinician's Perspective
[Review]
Abstract:
While ovarian tumors demonstrate considerable amount of biologic and histologic diversity, there is considerable overlap in their radiologic appearance, making their distinction difficult on radiologic or gross examination alone. The patient's age, family history, the presence of bilateral masses and the presence or absence of elevated tumor markers may narrow the differential diagnosis. In the following article, we will review the epidemiologic, clinical, and prognostic factors associated with the most common ovarian neoplasms.
Epithelial Ovarian Cancer
About 1 in every 57 women in the United States will develop epithelial ovarian cancer. Some studies have shown that breast-feeding and taking birth control pills may decrease a woman's likelihood of developing ovarian cancer. Moderate physical activity may also help women lower their chances of developing ovarian cancer, however, studies are conflicting. [1]
The most common type of ovarian cancer is the epithelial type (serous, mucinous, or endometrioid) that generally occurs in women in their late 50s and require distinction from their benign counterparts (serous or mucinous cystadenoma). Women who experience persistent lower abdominal discomfort or bloating should seek the advice of their physician and undergo a pelvic examination by a medical professional [2,3]. Women with suspected ovarian cancer should be referred to a gynecologic oncologist or a physician with special expertise in the management of women's cancers [4]. Prior to surgery, clinicians use diagnostic imaging and tumor markers to attempt to distinguish benign from malignant tumors. This is not only important for preoperative counseling but also to make certain that appropriate surgical expertise is available at the time of surgery.
Diagnostic Tests
An ultrasound should be considered to aid in the evaluation of a mass limited to the pelvis. Jacobs et al developed an index termed “risk of malignancy index” (RMI) that incorporates serum CA125, ultrasound findings (bilateral lesions, multilocular cystic lesions, solid areas, ascites, and intraabdominal spread), and menopausal status for the accurate preoperative diagnosis of ovarian cancer [5-7]. The RMI 2 score has a sensitivity of 74% to 80%, a specificity of 89% to 92%, and positive predictive values around 80%.
Patients with larger masses or with suspected ascites should have a computed tomography scan or magnetic resonance imaging study. Menopausal women should also have a CA125 blood test. CA125 is elevated in approximately 80% of cases of epithelial ovarian cancer. It is less likely to be elevated in cases of stage I disease or in patients with mucinous tumors. CA125 is a high-molecular-weight (>200 kD) glycoprotein antigenic determinant. It reacts with an antigen common to most nonmucinous epithelial ovarian carcinomas. It is positive in serous and endometrioid adenocarcinomas, Krukenberg tumors, and uterine leiomyomas. Mucinous tumors (benign, borderline, malignant) and normal ovarian epithelium are negative. The CA125 antigen is also detectable on the apical surface of normal mesoepithelial cells of the pleura, pericardium, and peritoneum and epithelium of the fallopian tube, endometrium, and endocervix. Normal breast, skin, liver, spleen, and kidney are negative. Other markers such as CEA may be considered if the patient predominantly presents with symptoms referable to the gastrointestinal tract, weight loss, pain, or hematochezia. If the primary is determined to be of gastrointestinal origin, the metastatic tumors in the ovaries may be removed for symptomatic relief, along with removal of the primary if appropriate.
Tumors of Low Malignant Potential (Borderline Tumors)
Ten percent of ovarian tumors are termed “tumors of low malignant potential” (LMP) or borderline tumors. While uncommon, these tumors are important to recognize in that they tend to occur in reproductive-age women who may be desirous of fertility or retention of the ovaries. These tumors may be treated by ovarian cystectomy or oophorectomy [8]. Patients that opt for ovarian cystectomy need to be aware of the risk of recurrence. Fortunately, most recurrences are amenable to surgical resection. Patients who have completed childbearing should undergo hysterectomy and removal of both tubes and ovaries. Performance of a lymph node dissection and omental resection for an LMP tumor grossly confined to the ovary may be considered, given the possible existence of invasive cancer not detected on frozen section, or the presence of occult disease in extraovarian tissue [9]. Distinguishing these tumors from an invasive cancer is critical to clinical decision making because patients with LMP tumors do not require treatment beyond surgery. A small subset of patients may have invasive so-called implants; these patients are treated as though they have invasive cancer. From a clinical perspective, the heterogeneous nature of mucinous tumors relative to that of serous tumors means that a mucinous tumor is more likely than a serous tumor to have invasive cancer identified on the permanent section that was not seen on the frozen. Thus, frozen-section reports should ideally attempt to convey the likelihood of the diagnosis changing on final review. Mucinous tumors may also represent metastases from a primary mucinous tumor of the appendix; thus, consideration should be given to performance of appendectomy at the time of surgery.
Surgical Considerations
Patients with invasive ovarian cancer should undergo cytoreductive surgery that typically involves removal of both ovaries and the uterus, as well as peritoneal biopsies and omentectomy. Maximal effort is made to resect as much gross tumor as possible because patients with minimal residual disease have improved prognosis when compared with patients who have bulky residual disease [10]. Patients with disease clinically confined to the ovary or pelvis should undergo pelvic and para-aortic lymph node sampling as occult disease may be present in up to 20% of apparent stage I cases.
In selected premenopausal patients with apparent stage I disease at the time of surgery, the contralateral ovary and uterus may be preserved. Although the majority of women with ovarian cancer require treatment with cytotoxic chemotherapy, carefully staged patients with stage I a-b, grade 1 disease may be safely observed without further therapy [11]. In these cases, the tumor stage, examination of lymph nodes, and assignment of tumor grade are critical to clinical decision-making.
Pathologic Considerations
Ovarian cancer is surgically staged; thus, both pathologic and clinical information is critical to assigning the appropriate stage. Important elements of reporting include grade, histologic type, sites of involvement, size of metastases, careful evaluation of submitted lymph nodes, and the presence or absence of malignant cytology.
Early Detection and Screening
The majority of women with ovarian carcinoma present with disease in the upper abdomen (Stage IIIc) at the time of diagnosis. Given that prognosis is linked with the extent of disease at the time of diagnosis, major research initiatives are under way to attempt to detect ovarian cancer at an earlier stage. Currently, there is no good screening test for ovarian cancer. The National Institutes of Health recently concluded a large screening trial that included more than 154,000 women and men between the ages of 55 and 74. Half of the women were randomized to receive CA125 and pelvic ultrasound and half received routine care by their health care providers. The patients are now being followed, and results will not be available for several years [12]. Although not recommended for normal risk premenopausal women, periodic ultrasound and examination of CA125 levels in the blood may be considered for patients with close family members who have ovarian cancer or women with mutations in a breast cancer susceptibility gene (BRCA1 or BRCA2) [13]. Carefully counseled patients with inherited genetic mutations may elect to undergo risk-reducing prophylactic oophorectomy. The ovaries of such patients may contain occult cancer and thus should be evaluated in their entirety.
Given the relatively low incidence of ovarian cancer, it has been a challenge to find a test that can identify early disease without falsely identifying too many women who do not have the disease. While a single assay of CA125 is not adequate for screening, investigators at Harvard Medical School have been looking at the value of measuring CA125 over time [14]. A statistical model converts the longitudinal CA125 profile to a single number called “risk of ovarian cancer,” which is the risk of having the disease at a given time. Studies suggest that longitudinal CA125 values increase sensitivity for ovarian cancer from about 70% to 85% while maintaining specificity. Other investigators have discovered new markers such as LPA or immunogenic ovarian tumor–associated antigens HOX A7 and HOX B7 (homeobox proteins) that have the potential to serve as biomarkers for early diagnosis of cancer [15-18]. Ardekani et al [19], Alexe et al [20], and Petricoin and colleagues [21] are attempting to identify proteomic patterns characteristic of ovarian cancer. To do this, they analyze proteins in serum with mass spectroscopy to look at varied subsets of proteins to detect any reproducible pattern of difference between normal and ovarian cancer. A computer algorithm that has been trained to recognize ovarian cancer–associated proteins is then employed to aid in distinguishing benign versus malignant tumors. Currently, investigators are refining their work and will soon open the world's first clinical reference laboratory for serum pattern proteomic diagnosis, which will handle serum samples for ovarian, breast, and prostate cancer in a clinical trials context. Initially NCI and the FDA will evaluate the combination of models in a clinical trial of high-risk women. It is anticipated that these studies will aid clinicians to define more precisely who is at risk for developing ovarian cancer.
Given the advanced stage at diagnosis for most women, treatment is not usually curative. A typical tumor will recur after primary therapy and require repeated courses of chemotherapy. The goal for health professionals must be to ensure that where cure is not possible, a woman can maintain good quality of life for as long as possible with judicious use of surgery and chemotherapy. Surgical and chemotherapy research questions need to be answered by large randomized controlled trials. Pathologists can aid in this endeavor through continued research in the emerging areas of molecular pathology, timely and detailed reporting of results, and submission of tissue for patients entering clinical trials.
Germ Cell Tumors Of The Ovary
Germ cell tumors are diverse in their pathologic, as well as physiologic, characteristics and account for 15%–20% of all ovarian neoplasms. These tumors occur in younger women, with an average age of 19 years. Although 60% to 75% of these tumors are stage I at diagnosis, they can grow rapidly and metastasize widely [22]. Successful treatment of these tumors is dependent on correct pathologic classification since different chemotherapeutic agents and surgical approaches are used, depending on tumor histology. Patients with germ cell tumors should also be evaluated for possible gonadal dysgenesis since gonadoblastomas can develop into germ cell tumors.
Many germ cell tumors are highly curable with appropriate treatment, and therefore referral to a gynecologic oncologist for evaluation and treatment is of paramount importance for these young women. Because most of these tumors are limited to one ovary, unilateral salpingo-oophorectomy and a careful staging procedure with reproductive conservation is standard treatment. A germ cell tumor is not always easy to diagnose on frozen-section analysis. Histologically, some germ cell tumors can resemble poorly differentiated adenocarcinomas, clear cell carcinoma, lymphoma, or other malignancies [23]. In young women with a possible germ cell tumor encountered at surgery, the patient is best served with a more conservative procedure rather than a hysterectomy and bilateral salpingo-oophorectomy. In this way, many young women who suffer from these cancers can go on to mature normally and remain fertile.
Benign Germ Cell Tumors
The mature cystic teratoma or dermoid cyst is the most common germ cell tumor of the ovary. These tumors are the most common type of ovarian neoplasm in childhood and in women under 35. These tumors are bilateral in about 15% of cases and in most cases are treated adequately by ovarian cystectomy. Malignancy occurs in 1%–2% of dermoid cysts, and therefore all dermoid cysts deserve careful histologic examination. Most of the malignancies that arise in benign teratomas are seen in women over 50 and are usually squamous carcinomas 24]. These tumors are treated by excision of the involved adnexa with careful peritoneal exploration and possible adjuvant therapy [23].
Malignant Germ Cell Tumors
Dysgerminoma
Dysgerminoma is a very curable germ cell tumor of the ovary that most commonly presents between the ages of 10 and 30. Histologically, these tumors resemble the testis. Dysgerminomas make up 40% of all germ cell malignancies and are unusual in that they have a significant rate (15%) of bilaterality [22]. Dysgerminomas are also marked by their propensity for lymphatic, rather than intraperitoneal, spread. Therefore, the surgical treatment of these tumors should always include lymph node biopsies, as well as a thin biopsy of the opposite ovary. Dysgerminomas are the most common ovarian cancers to arise from gonadoblastomas, and therefore possible chromosomal abnormalities must be excluded.
Endodermal Sinus Tumor
Endodermal sinus, or yolk sac, tumors are highly malignant and present at a median age of 18 years. One third of patients are prepubertal at the time of diagnosis and frequently present complaining of abdominal pain or mass [25]. Most of these tumors secrete [alpha] fetoprotein, which can be used to monitor therapy. All patients with endodermal sinus tumors receive chemotherapy, and most tumors will respond to treatment. Schiller Duval bodies are a classic feature of endodermal sinus tumors. In some cases, endodermal sinus tumors can be difficult to distinguish from a clear-cell carcinoma or high-grade adenocarcinoma [23].
Immature Teratomas
The immature teratoma is a malignant germ cell tumor marked by the presence of immature tissues similar to those seen in the human embryo. Areas of mature cystic teratoma are often found within these tumors as well and can mislead the pathologist if the immature components are not noted.
Once the diagnosis of immature teratoma is made, whether or not the patient requires adjuvant therapy is dependent on the stage at diagnosis, as well as histologic grade. Other malignant germ cell types such as endodermal sinus tumor can be present within these tumors as well and should be noted since they will guide therapy.
Immature teratomas are generally large tumors and can present with peritoneal implants and ascites. Interestingly, these implants can consist of nests of mature glia, which does not portend a worse prognosis [26].
Other Germ Cell Tumors
Polyembryoma is a very rare ovarian tumor that presents at a median age of 14 years. These tumors may secrete estrogens, leading to abnormal bleeding. Both polyembryomas and primary choriocarcinomas are rare tumors of the ovary and can cause precocious puberty in children.
Generation of Pathology Reports for Germ Cell Tumors of the Ovary
Generation of an inclusive pathology report is critical to formulation of an appropriate plan of care for the patient with a germ cell tumor of the ovary. What type of surgery, chemotherapy, or radiation is used for treatment varies widely between different tumors. Therefore, the clinician is dependent on the histologic diagnosis and, in most tumors, surgical stage. There are numerous other prognostic factors that are important to report in different tumors. For dysgerminomas, the presence of advanced stage, size over 15 cm, a medullary pattern, and numerous mitoses are all worrisome features associated with an increased risk of recurrence [22,27]. In the immature teratoma, the most important prognostic factor is grade of the tumor. The prognosis of mixed germ cell malignancies depends on tumor composition and what percentage of the tumor is made up of its most malignant component [28]. Through accurate identification and reporting of germ cell tumors, the pathologist can make a vital difference in the care of these women with cancer.
Sex-Cord Stromal Tumors
Sex-cord stromal tumors are composed of sex-cord and stromal elements that are of ovarian (female) differentiation (granulosa and theca cells) or testicular (male) differentiation (Sertoli and Leydig cells), both alone and in combination, as well as stromal fibroblasts and indifferent cells. These tumors are classified according to the World Health Organization and account for the majority of functioning ovarian tumors and 6% of malignant ovarian neoplasms [29,30]. These tumors are fairly rare, have quite an indolent natural history, and overall are associated with a favorable prognosis.
Granulosa Cell Tumors
Approximately 70% of sex-cord stromal tumors are granulosa cell tumors (GCTs), though they account for less than 2% of all ovarian tumors and 3%–5% of ovarian malignancies [31]. There are 2 subtypes, the adult GCT (95%) and the juvenile GCT. Most of the juvenile GCTs (less than 5% of all GCTs) appear prior to puberty, while one third of all GCTs present in women of reproductive age and the remainder in postmenopausal women [32]. Over 70% of these tumors secrete estrogen, and over 60% of patients present with related symptoms of abnormal uterine bleeding [33]. Reproductive-age women typically encounter menometrorrhagia, while postmenopausal women may present with bleeding. Other symptoms may include abdominal pain, increased abdominal girth, a palpable mass, hemoperitoneum from rupture, or rarely, signs of hyperandrogenism [34]. GCTs are usually unilateral and are comprised of both cystic and solid components, with the solid component appearing yellow and the cysts containing blood, but the appearance can vary from a completely solid tumor to unilocular or multilocular cystic tumors. They average 12 cm in diameter, with 27% over 15 cm, 24% between 11 and 15 cm, 38% from 6 to 10 cm, and 11% under 6 cm. Inhibin, a gonadal peptide produced by granulosa cells, can be useful to the clinician as a serum tumor marker in assessing recurrence and response to treatment [35,36]. Anti-Mullerian Hormone and serum estradiol may also serve as tumor markers [37].
As with most sex-cord stromal tumors and epithelial tumors of LMP, the primary treatment of GCTs is surgical, with resection of all visible disease, as well as staging. It is also important to evaluate the endometrium, as there is a 25%–50% risk of endometrial hyperplasia and a 5%–10% risk of concurrent endometrial carcinoma. Pelvic and paraortic lymphadenectomy is indicated for staging [38]. As GCTs are bilateral in just 2% of cases, a unilateral salpingo-oophorectomy is acceptable for stage Ia tumors in premenopausal women. No adjuvant treatment is recommended for women with early-stage disease. Given the potential for late recurrence, surveillance is usually continued for at least 10 years [39]. In those patients with advanced disease and macroscopic residual disease, adjuvant therapy is generally employed, though the optimal regimen has not been determined [40,41].
Thecomas and Fibromas
Thecomas and fibromas are benign stromal cell tumors. Like GCTs, thecomas can produce estrogen, commonly present with postmenopausal bleeding, and may be associated with endometrial hyperplasia or endometrial carcinoma. While removal of the affected ovary and staging should be performed in premenopausal women, total abdominal hysterectomy, bilateral salpingoophorectomy, and lymphadenectomy should be performed in postmenopausal women. Coexistent endometrial pathology should be excluded in all patients. These tumors may occasionally be associated with Meigs syndrome. The ascites and pleural effusions seen with Meigs syndrome are likely caused by cytokines that increase capillary permeability and can result in an elevated preoperative CA125 level [42]. The fluid resolves following removal of the ovarian tumor. Fibromas may also be associated with Gorlin syndrome, an autosomal-dominant basal nevus cell syndrome.
Sertoli-Leydig Cell Tumor
Sertoli-Leydig cell tumor (SLCTs) are uncommon, accounting for less than 0.5% of ovarian tumors; they generally occur in younger patients, with an average age at diagnosis of 25, but can be seen at all ages [43]. Sertoli cell tumors are even more rare. SLCTs present with hirsutism or virilization in half of the patients; if endocrine symptoms do not manifest, then patients present with symptoms of a pelvic mass. Grossly these tumors are solid and yellow-tan due to the lipid content and can contain cystic areas. The degree of differentiation is important in prognosis, and fortunately most are low grade. Other factors of prognostic significance are the stage, the presence of heterologous elements, mitotic index, and tumor rupture. Sertoli-Leydig tumors can produce large amounts of testosterone, androstenedione, or both. Conservation of the contralateral ovary may be considered for women desirous of preservation of fertility. Surgical treatment may include removal of the affected ovary and surgical staging for premenopausal women and total abdominal hysterectomy, bilateral salpingo-oophorectomy, and surgical staging for postmenopausal women. Adjuvant therapy is considered for all poorly differentiated tumors, those with heterologous elements, or advanced-stage disease [44]. Pathologically, these tumors can be confused with endometrioid carcinomas, tubular Krukenberg tumors, carcinoid tumors, and endodermal sinus tumors, particularly in pregnant patients.
REFERENCES
1. Pan S, Ugnat A, Mao Y. Physical activity and the risk of ovarian cancer: a case-control study in Canada. Int J Cancer. 2005;117:300–307.
2. Goff B, Mandel L, Melancon C, et al. Frequency of symptoms of ovarian cancer in women presenting to primary care clinics. JAMA. 2004;291:2705–2712.
3. Brooks SE, Langenberg P, Baquet CR, et al. Early detection and screening for ovarian cancer: does physician specialty matter? J Assoc Acad Minor Phys. 2002;13:69–75.
4. Im S, Gordon A, Buttin B, et al. Validation of referral guidelines for women with pelvic masses. Obstet Gynecol. 2005;105:35–41.
5. Jacobs IJ, Skates SJ, MacDonald N, et al. Screening for ovarian cancer: a pilot randomised controlled trial. Lancet. 1999;353:1207–1210.
6. Menon U, Talaat A, Jeyarajah AR, et al. Ultrasound assessment of ovarian cancer risk in postmenopausal women with CA125 elevation. Br J Cancer. 1999;80:1644–1647.
7. Jacobs I, Oram D, Fairbanks J, et al. A risk of malignancy index incorporating CA125, ultrasound and menopausal status for the accurate preoperative diagnosis of ovarian cancer. Br J Obstet Gynaecol. 1990;97:922–929.
8. Chambers JT. Borderline ovarian tumors: a review of treatment. Yale J Biol Med. 1989;62:351–365.
9. Morice P, Leblanc E, Rey A, et al. Conservative treatment in epithelial ovarian cancer: results of a multicentre study of the GCCLCC (Groupe des Chirurgiens de Centre de Lutte Contre le Cancer) and SFOG (Societe Francaise d'Oncologie Gynecologique). Hum Reprod. 2005;20:1379–1385.
10. Hoskins WJ. Epithelial ovarian carcinoma: principles of primary surgery. Gynecol Oncol. 1994;55(3 pt 2):S91–96.
11. Colombo N, Guthrie D, Chiari S, et al. International Collaborative Ovarian Neoplasm Trial 1: a randomized trial of adjuvant chemotherapy in women with early-stage ovarian cancer. Natl Cancer Inst. 2003;95:125–132.
12. Hartge P, Hayes R, Sherman M, et al. Complex ovarian cysts in postmenopausal women are not associated with ovarian cancer risk factors: preliminary data from the PLCO cancer screening trial. Ann Epidemiol. 2000;10:465.
13. Coupier I, Gauthier-Villars M, This P, et al. [Genetic predisposition and ovarian cancer]. Rev Prat. 2004;54:1757–1762.
14. Skates SJ, Menon U, MacDonald N, et al. Calculation of the risk of ovarian cancer from serial CA-125 values for preclinical detection in postmenopausal women. J Clin Oncol. 2003;21(10 suppl):206–210.
15. Xiao Y, Chen Y, Kennedy AW, et al. Evaluation of plasma lysophospholipids for diagnostic significance using electrospray ionization mass spectrometry (ESI-MS) analyses. Ann N Y Acad Sci. 2000;905:242–259.
16. Naora H, Montz FJ, Chai CY, et al. Aberrant expression of homeobox gene HOXA7 is associated with mullerian-like differentiation of epithelial ovarian tumors and the generation of a specific autologous antibody response. Proc Natl Acad Sci U S A. 2001;98:15209–15214.
17. Xu Y, Shen Z, Wiper DW, et al. Lysophosphatidic acid as a potential biomarker for ovarian and other gynecologic cancers. JAMA. 1998;280:719–723.
18. Sutphen R, Xu Y, Wilbanks GD, et al. Lysophospholipids are potential biomarkers of ovarian cancer. Cancer Epidemiol Biomarkers Prev. 2004;13:1185–1191.
19. Ardekani AM, Liotta LA, Petricoin EF 3rd. Clinical potential of proteomics in the diagnosis of ovarian cancer. Expert Rev Mol Diagn. 2002;2:312–320.
20. Alexe G, Alexe S, Liotta L, et al. Ovarian cancer detection by logical analysis of proteomic data. Proteomics. 2004;4:766–783.
21. Petricoin EF, Ardekani AM, Hitt BA, et al. Use of proteomic patterns in serum to identify ovarian cancer. Lancet. 2002;359:572–577.
22. Kurman R, Norris H. Malignant germ cell tumors of the ovary. Human Pathol. 1997;8:551.
23. Morrow C. Tumors of the Ovary: Sex Cord Stromal and Germ Cell. Vol 281. 5th ed. Philadelphia: Churchill Livingstone; 1998.
24. Gordon A, Rosenshein N, Parmley T, et al. Benign cystic teratomas in postmenopausal women. Am J Obstet Gynecol. 1980;138:1120.
25. Imai A, Furui T, Tamaya T. Gynecologic tumors and symptoms in childhood and adolescence: ten year's experience. Int J Gynaecol Obstet. 1994;45:227–234.
26. Truong LD, Jurco S III, McGavran M. Gliomatosis peritonei: report of two cases and review of literature. Am J Surg Pathol. 1982;6:443.
27. Kurman RJ, Scardino PT, Waldmann TA, et al. Malignant germ cell tumors of the ovary and testis: an immunohistologic study of 69 cases. Ann Clin Lab Sci. 1979;9:462–466.
28. Gershenson D, Del Junco G, Copeland L, et al. Mixed germ cell tumors of the ovary. Obstet Gynecol. 1984;64:200–206.
29. Young RH, Scully RE. Ovarian sex cord-stromal tumors: recent advances and current status. Clin Obstet Gynaecol. 1984;11:93–134.
30. Serov S, Scully R, Sobin L. Histologic typing of ovarian tumors. In: World Health Organization: International Histological Classification of Tumors. Geneva; 1973.
31. Novak E. Review of the granulose-theca cell tumors from the Emil Novak Ovarian Tumor Registry. Am J Obstet Gynecol. 1999;10:323.
32. Scully RNS. Tumors of the Ovary and Maldeveloped Gonads. Vol 32–33. Washington, DC: Armed Forces Institute of Pathology; 1979:32–33, 152–225.
33. Schwartz P, Smith J. Treatment of ovarian stromal tumors. Am J Obstet Gynecol. 1976;125:402.
34. Ohel G, Kaneti H, Schenker J. Granulosa cell tumors in Israel: a study of 172 cases. Gynecol Oncol. 1983;15:278.
35. Hildebrandt RH, Rouse RV, Longacre TA. Value of inhibin in the identification of granulosa cell tumors of the ovary. Hum Pathol. 1997;28:1387–1395.
36. Robertson DM, Burger HG, Fuller PJ. Inhibin/activin and ovarian cancer. Endocr Relat Cancer. 2004;11:35–49.
37. Rey RA, Lhomme C, Marcillac I, et al. Antimullerian hormone as a serum marker of granulosa cell tumors of the ovary: comparative study with serum alpha-inhibin and estradiol. Am J Obstet Gynecol. 1996;174:958–965.
38. Schumer ST, Cannistra SA. Granulosa cell tumor of the ovary. J Clin Oncol. 2003;21:1180–1189.
39. Lauszus FF, Petersen AC, Greisen J, et al. Granulosa cell tumor of the ovary: a population-based study of 37 women with stage I disease. Gynecol Oncol. 2001;81:456–460.
40. Uygun K, Aydiner A, Saip P, et al. Clinical parameters and treatment results in recurrent granulosa cell tumor of the ovary. Gynecol Oncol. 2003;88:400–403.
41. Homesley HD, Bundy BN, Hurteau JA, et al. Bleomycin, etoposide, and cisplatin combination therapy of ovarian granulosa cell tumors and other stromal malignancies: a Gynecologic Oncology Group study. Gynecol Oncol. 1999;72:131–137.
42. Abramov Y, Anteby SO, Fasouliotis SJ, et al. Markedly elevated levels of vascular endothelial growth factor, fibroblast growth factor, and interleukin 6 in Meigs syndrome. Am J Obstet Gynecol. 2001;184:354–355.
43. Young RH, Scully RE. Ovarian Sertoli-Leydig cell tumors: a clinicopathological analysis of 207 cases. Am J Surg Pathol. 1985;9:543–569.
44. Zaloudek C, Norris HJ. Sertoli-Leydig tumors of the ovary: a clinicopathologic study of 64 intermediate and poorly differentiated neoplasms. Am J Surg Pathol. 1984;8:405–418.
45. Pecorelli S, Benedet JL, Creasman WT, et al. FIGO staging of gynecologic cancer: 1994–1997 FIGO Committee on Gynecologic Oncology: International Federation of Gynecology and Obstetrics. Int J Gynaecol Obstet. 1999;65:243–249.
Pathology Reviews, Volume 11(1), January/February 2006, pp 3-8
Borderline Ovarian Tumors: Consensus, Controversy, and Continuing Challenges
[Review]
Abstract:
The subject of borderline ovarian tumors (BOTs) has incited numerous controversies, including but not limited to questions of nomenclature, definitions, diagnosis, and treatment. A workshop was held in August 2003, under the auspices of the National Cancer Institute, to discuss these controversies and attempt to resolve at least some of them. The present article discusses the controversial subjects under the headings of Nomenclature, Quantitative Aspects (of definitions, sampling, and diagnosis), Problems and Controversies Limited to Serous BOTs (including the micropapillary variant, extraovarian “implants,” and intranodal epithelial lesions), Problems and Controversies Unique to Mucinous BOTs (including intraepithelial carcinoma, metastatic mucinous carcinoma in the ovary, and pseudomyxoma peritonei); and Seromucinous/Endocervical/Mucinous Müllerian BOTs. Recommendations of the workshop, personal prejudices of the author, and suggestions for the practicing pathologist dealing with these tumors are presented.
Since their original description by Taylor [1] as “semimalignant tumors of the ovary,” ovarian borderline tumors have continued to incite controversy, particularly since they became more commonly recognized about 30 years ago. In August 2003, the Borderline Ovarian Tumor (BOT) Workshop was held in Bethesda, MD, under the auspices of the National Cancer Institute. The purpose of this workshop, which was organized by Dr. Jules J. Berman and me, was to bring together a group of physicians (predominantly pathologists) who have been interested in this subject to at least define—and with luck, at least partially resolve—the controversies that have arisen in the nomenclature, definition, diagnosis, and management of these tumors and to suggest pathways for future research. A summary of the proceedings of this workshop was published in August 2004, and the reader is referred to the publications at that time for a more detailed discussion of the issues, which will be reviewed briefly in the current article [2-7]. As with the BOT workshop, this article will be limited to serous and mucinous BOTs, which comprise the great majority of these tumors. Although it was generally agreed at the workshop that endometrioid, clear cell, and transitional cell (Brenner) borderline tumors were deserving of discussion and an attempt at consensus, time ran out before a thorough discussion of these lesions could take place. It is my personal opinion, however, that most of the lesions diagnosed and recorded as endometrioid BOTs are probably in reality low-grade endometrioid carcinomas, with the remainder being either atypical endometrioid adenofibromas or atypical endometriosis; that most putative clear cell BOTs are benign atypical clear-cell adenofibromas; and that borderline Brenner tumors do exist but are extremely rare, with no clear evidence of malignant behavior in the literature to date.
Nomenclature
Although the original terminology of “semimalignant” used by Taylor [1] is no longer accepted, it has been replaced by not 1 but 3 terms, which are in current use. In the World Health Organization (WHO) classification system of 1973 [8], these tumors were referred to as “tumors of borderline malignancy (carcinomas of low malignant potential).” The most recent WHO classification of 2003, however, designated “borderline tumor” as the preferred term and “tumor of low malignant potential” as an accepted synonym [9], and “borderline tumor” was also the choice in the most recent Armed Forces Institute of Pathology (AFIP) Atlas of Tumor Pathology fascicle on ovarian tumors. Finally, the term “atypical proliferative tumor” was first used by Dr. Peter Russell in the 1970s and continues to be used by him [11] and by Seidman and Kurman [12] to describe the majority of these lesions.
At the BOT workshop, these 3 designations all had their adherents, and the only agreement reached was that the terms are meant to be and should be considered synonymous. It was also agreed almost unanimously that these tumors should not be called carcinomas of low malignant potential and that one of the most important roles of the pathologist is to distinguish these tumors from true invasive ovarian carcinomas.
For the practicing pathologist, the most important considerations concerning these terminological disputes are the following. First is the recognition that published articles and monographs on ovarian tumors have used and will continue to use these 3 terms more or less synonymously, and second is the recommendation that all pathologists within a single group practice choose and use a single term so as not to confuse their clinicians. My personal recommendation, as made clear by the title of this article, is the term “borderline tumor” since this is currently preferred by the WHO.
Additionally, it should be remembered that, although serous and mucinous BOTs are discussed together in this article and most other publications, significant differences exist between these histopathologic types at the epidemiologic, molecular, diagnostic, and management levels [13-16]. Thus, parts of the discussion below will emphasize commonalities between the different histopathologic types, while other parts will deal with issues specific to either serous or mucinous BOTs.
In this connection, it should be noted that the BOT workshop recommendations included the statement that “publications on borderline tumors should completely separate the data by cell type (eg, serous, mucinous).” It was also recommended that publications report the degree of sampling in all cases; include specific per-case data for those with an adverse outcome; report how the adverse outcome was documented (biopsy, autopsy, other method), with a histologic description of what was found; describe both the ovarian tumor(s) and any extraovarian disease; and provide more than brief follow-up data because of the natural history of these tumors. It was also recommended that the histopathologic details be reviewed in both clinical studies and nonmorphological basic investigative studies by a diagnostic pathologist who is an active member of the research team [3].
Quantitative Aspects
Although the general definition of a BOT is that of a noninvasive epithelial ovarian neoplasm with stratification, nuclear abnormalities, and mitotic activity intermediate between benign and malignant tumors of similar cell type [4,14], certain quantitative characteristics of the upper and lower boundaries of this definition have never been defined, nor has the extent of sampling required to verify the diagnosis. With regard to the latter, the BOT workshop recommended the submission of at least 1 section per centimeter of maximal tumor dimension for tumors smaller than 10 cm and 2 sections per cm for larger tumors (excluding smooth-walled cystic foci). The commentary accompanying this recommendation, however, stated that “It is not at all clear that 10 cm is a logical point at which to change the sampling technique. An alternative, more cost-effective system would be to increase the number of sections only if a worrisome finding (eg, intraepithelial carcinoma, microinvasion) were identified in the original sections. Prospective studies to develop an evidence-based rationale for sampling of both primary tumors and omentectomy specimens are clearly in order.” [3]
With respect to gross description and sampling, it was also recommended that the former include the presence of an exophytic or intracystic growth pattern, as well as capsular rupture, and that in the presence of capsular rupture, sections be taken from the immediate vicinity (often in collaboration with the gynecologic oncologist who performed the surgery) and be identified as such [3].
In addition to the extent of sampling necessary to rule in or out a specific feature, the proportion of a tumor demonstrating the characteristic morphologic appearance of a BOT, as well as proportions necessary for the diagnosis of intraepithelial carcinoma (in the case of mucinous BOTs), micropapillary variant (of serous BOT), microinvasion, and invasive carcinoma are also not universally agreed upon. The BOT workshop pointed out that no definitive studies are available on the significance of small foci of tumor with borderline morphology within an otherwise benign ovarian tumor (cystadenoma, adenofibroma, etc.), and recommended “If a benign element is present, then the percentage or greatest measurement of the borderline component should be mentioned in the pathology report.” Most (but not all) participants indicated that they would be willing to diagnose BOT if 10% or more of sampled tumor showed the typical morphology, but it was (and remains) unclear whether smaller foci should be mentioned or ignored. Although the BOT workshop called for more studies to resolve this conundrum, it is extremely unlikely that the appropriate study will ever be done, considering the favorable survival data for even tumors with 100% borderline morphology. Similarly, whether the proportion required for a borderline diagnosis should differ between serous and mucinous tumors is unknown and probably unknowable.
At the opposite end of the spectrum of the diagnosis of BOT is its distinction from microinvasive and fully invasive carcinoma. Here again, although there is probably a majority viewpoint, many different definitions for these entities have been proposed in the literature [4]. With respect to microinvasion, the major differences in criteria involve (1) the appearance of the infiltrating tumor cells and (2) the maximum dimension(s) of an invasive focus that can still be classified as microinvasion rather than invasive serous or mucinous carcinoma.
Since microinvasive foci may vary from a few single cells within cleftlike spaces or a minimally reactive stroma to small foci satisfying size criteria for microinvasion but architecturally resembling invasive carcinoma, it is probably not surprising that some authors have used different terms for these 2 patterns: “BOT with microinvasion” for the former and “BOT with microinvasive carcinoma” for the latter. In addition, in mucinous intestinal BOTs, the term “mucinous borderline tumor with microinvasive carcinoma” was reserved for cases in which the noninvasive component contained intraepithelial carcinoma (see below) [17].
Variability in size criteria exists as well, the definitions of the upper limit of microinvasive foci having been quoted as 3 mm and 5 mm in greatest linear dimension and 10 mm3 in 2 dimensions. Because of limited experience in the literature with this group of tumors, there was no consensus among the participants of the BOT workshop for a size criterion, but it was noted that the most commonly accepted criterion is up to 5 mm of greatest linear measurement in any single focus [3]. It should also be noted that multiple foci of microinvasion often occur in a single case, and these do not change the diagnosis (or the prognosis, which is uniformly excellent) if all foci noted conform to the size criterion used. Whether or not the cytologic features (particularly high-grade versus low-grade) of the microinvasive foci correlate with any difference in prognosis is currently unknown. It should be noted that, although some attention has been given here and elsewhere in the literature to the maximum dimension(s) permissible for the diagnosis of microinvasion, the great majority of foci seen in practice are considerably smaller than any of these. It is also notable that, in our own series of microinvasive BOTs, microinvasion was noted in approximately 10% of all BOTs (serous, mucinous intestinal, and mucinous müllerian) but had been missed in the original pathologic interpretation of every case [18]. Thus, this is a lesion likely to be underdiagnosed in routine practice, and fortunately the lack of prognostic or therapeutic significance does not carry a penalty for either the patient or the pathologist when the diagnosis is missed.
The upper size limits for BOT with microinvasion, as defined above, obviously also define the lower size limits for fully invasive ovarian carcinoma. Although the BOT workshop was limited to problems concerning borderline tumors and thus did not comment to any great extent on invasive carcinomas, it seems intuitive that an invasive carcinoma diagnosed on the basis of a single 6-mm focus of what would otherwise be a BOT will probably be less aggressive than an invasive carcinoma completely replacing an ovary which measures 10 cm in greatest dimension. Thus, when invasive carcinoma is diagnosed on the basis of minimal criteria, a comment to this effect should probably be made in the surgical pathology report.
It should also be noted at this point that there remains some controversy regarding the diagnosis of invasive carcinoma in 2 situations. The first of these is the lesion referred to by an increasing number of authors as low-grade serous carcinoma, which must be distinguished from the micropapillary variant of serous BOT (to which it is probably etiologically related) on the one hand and the far more frequent high-grade serous carcinoma on the other [12,19]. The second relatively new and perhaps still controversial area concerns the expansion of the definition of invasive mucinous carcinoma to include an “expansile” or “confluent glandular” pattern of invasion, in which the glands are markedly crowded but do not induce a reactive desmoplastic stroma [6,20]. Since adverse prognostic implications are almost exclusively associated with the classic infiltrative rather than the expansile type of mucinous adenocarcinoma, I believe that many of these latter tumors should be considered borderline tumors or (at worst) borderline tumors with intraepithelial carcinoma.
Problems/Controversies Unique to Serous Borderline Tumors
Although the questions raised thus far in this article apply equally to serous and mucinous borderline tumors, there are also controversial issues specific to one or the other of these. Problems related specifically to serous BOTs include (1) the micropapillary variant (alternatively termed “noninvasive micropapillary carcinoma”), (2) the definition and diagnosis of extraovarian “implants,” and (3) lesions found in lymph nodes of patients with serous BOTs. These will be discussed in sequence here.
Micropapillary Variant of Serous BOT
One of the classic tenets for the diagnosis of borderline tumors in general has always been that they lack stromal invasion, whereas the corresponding carcinomas are by definition invasive. In 1996, Burks and colleagues [21] amended this formula by proposing the term “micropapillary serous carcinoma of the ovary” for a noninvasive lesion with distinctive histologic features. In this and subsequent publications, these authors and their colleagues (summarized in Silverberg et al [3], Seidman et al [4], and Bell et al [5]) have argued that the distinctive growth pattern of this tumor variant, when encountered within a typical serous BOT or by itself, should be diagnosed as carcinoma because tumors of this sort have higher frequencies (compared with typical serous BOTs) of bilaterality, surface ovarian involvement, advanced stage (manifested by extraovarian disease), and “invasive implants.” Most of these observations are probably correct and certainly suggest at the very least that women with this micropapillary tumor should always have a careful and extended staging laparotomy to search for invasive disease outside the ovary. On the other hand, the adverse effect of the micropapillary pattern on both recurrence rates and long-term survival seems to be limited to the cases in which invasive disease is found outside the ovary. In consideration of these data, the BOT workshop stated, “micropapillary serous borderline tumors (defined as serous borderline tumors exhibiting at least 1 5-mm focus of micropapillary and/or cribriform growth) should be distinguished from typical serous borderline tumors in the pathology report. The significance of tumors exhibiting smaller foci of micropapillary morphology needs further study. When such foci are encountered, further sampling is encouraged. As with other serous borderline tumors, the pathology report should distinguish between intracystic and exophytic tumor and more research is needed on the clinical implications, if any, of this distinction.” [3,22,23] Thus, the majority opinion at the present time is not to call these lesions “carcinoma” unless they are invasive. Those that do become invasive within the ovary (and in extraovarian sites as well) tend to do so in a distinctive pattern of low-grade serous carcinoma, usually characterized by small glands, solid nests, and micropapillae lined by cells with mild to moderate atypia and often surrounded by slitlike spaces. This pattern occurs in about 10% of invasive serous carcinomas and is usually easy to distinguish from the classic pattern, characterized by larger papillae and solid nests and much more marked nuclear atypia. Clinical, as well as molecular, evidence suggests that this distinction is important and that the low-grade variant of serous carcinoma probably arises from micropapillary serous borderline tumor, while the origin of the more common high-grade variant is at the present time unknown [3-5,12,13,19,22,24,25].
Finally, although this variant of serous BOT is designated “micropapillary,” it is now generally agreed that a cribriform pattern is also characteristic of this tumor. The micropapillary pattern is characterized by a nonhierarchical (nonbranching) growth of long, slender, hairlike micropapillae extending outward from large-caliber fibrovascular cores and lined by relatively uniform cells with grade 1 to 2 nuclear features. This pattern of growth often coexists with, and occasionally is completely replaced by, the cribriform pattern [4].
Peritoneal “Implants” and Intranodal Epithelium in Women With Serous BOTs
Although it is generally agreed that serous-BOT patients with extraovarian peritoneal “invasive implants” are at greater risk for recurrence and death—often from disseminated invasive carcinoma—both “invasive” and “implants” are terms the meaning of which is not totally clear. Although the use of the term “implants” implies that these arise by breaking off from surface ovarian epithelial tumor and subsequently implanting and growing elsewhere on peritoneal surfaces, it is far from clear that this is indeed the pathogenesis of these lesions in all cases or even in the majority of cases. It seems to be true that extraovarian “implants” are associated more frequently with exophytic than with intracystic serous BOTs, whether of the classic or the micropapillary type [22,23]. This suggests that at least some of these “implants” really do develop by the mechanism discussed above, but the frequency of multifocal primary disease accounting for the extraovarian lesions is not defined, and molecular studies have been confusing, indicating both monoclonal and polyclonal origins in different tumors and in different studies [7, 26-29].
The definition of these “implants”—if so they be—as invasive versus noninvasive is also somewhat controversial, as reflected in the fact that, although most reported series have documented poorer survival in patients with “invasive implants,” this has not been a universal experience [3,5,12,30-32]. This discrepancy is based largely on the application of different criteria for invasion by different authors. The most widely accepted criteria in recent years have been those of D. Bell and colleagues [33], who use the term “invasive” to refer to invasion of underlying tissue (omentum, bowel wall, etc.). Those lesions that appear to be “stuck on” the peritoneal surfaces, without invasion of underlying tissue, are characterized as noninvasive and divided into those that exhibit epithelial proliferation without a stromal response (epithelial type) and those that are accompanied by a marked stromal reaction (desmoplastic type). It is particularly important for the pathologist to recognize the latter type of noninvasive lesion since a desmoplastic stromal response in other situations is usually regarded as evidence of invasive cancer. It is useful to recognize that, in most desmoplastic “implants,” the desmoplastic stromal reaction comprises at least 75% (and often considerably more) of the total mass.
Although other authors have largely accepted this distinction between noninvasive and invasive extraovarian lesions, several have noted that, in the situation in which only small surface biopsies are submitted to the pathologist, it is often impossible to determine the invasive versus noninvasive interface between the lesion and underlying normal tissues. This problem has led to the presentation by K. Bell and colleagues [34] of a set of “refined diagnostic criteria” for invasive versus noninvasive implants, which, in addition to tissue invasion, include a micropapillary architecture and solid epithelial nests or papillae surrounded by a cleft [4,5]. At the present time, the prognostic validity of these new criteria has not been confirmed by other large studies, but it has always appeared to me that invasive carcinoma should be diagnosable as such even on a small superficial biopsy. In many if not most instances, of course, the diagnosis of peritoneal lesions by the D. Bell and the K. Bell criteria will be consonant, but it remains to be determined how to classify those lesions in which the diagnosis would be noninvasive by one system and invasive by the other. This is of more than academic interest since most gynecologic oncologists will offer chemotherapy to women diagnosed with “invasive implants” and may or may not withhold it from women whose extraovarian lesions are called noninvasive.
Although space limitations preclude an extensive discussion in this review of peritoneal lesions considered to be less than “implants,” it is important to note that endosalpingiosis (which is considered a benign lesion) is seen in many women with serous BOTs and is considered benign, thus not converting the stage of the ovarian tumor upward. I would personally interpret as atypical endosalpingiosis some lesions that have been reported as noninvasive epithelial-type serous BOT implants and would point out that the distinction between these 2 lesions—both of which seem to have a favorable prognosis but one of which upstages the ovarian tumor—needs further investigation and consensus. Parenthetically, although all gynecologic pathologists regard endosalpingiosis as a benign lesion, Silva and colleagues [35] have noted a significantly higher frequency of endosalpingiosis in extraovarian sites in women with stage I serous BOT who experienced recurrence than in a control group of women with no recurrence after a minimum follow-up of 15 years. Moore et al [36] also suggested that similar lesions in lymph nodes might actually sometimes be metastases from serous BOTs and have noted identical K-ras mutations in some of these lymph node lesions and in the associated serous BOTs [37]. Thus, the significance of these lesions associated with serous BOTs may need to be reconsidered in the future.
In addition to presumptively benign glandular inclusions, other lymph node lesions have been reported in women with serous BOTs. Sites of involvement, in decreasing order of frequency, have included paraortic, pelvic, iliac, obturator, and omental lymph nodes [5]. The lymph node lesions usually consist of either individual cells with abundant eosinophilic cytoplasm or papillary clusters of serous BOT-like cells, although both may coexist in the same node and usually also coexist with müllerian inclusions or endosalpingiosis. Although the usual assumption (and certainly my own) has been that these lymph node lesions arise from the benign inclusions in the same way that ovarian serous BOTs arise from surface and/or inclusion cyst epithelium, the studies of Moore et al [36] and Alvarez et al [37], as mentioned above, suggest the possibility of the differentiation taking place in the opposite direction. In any event, death has not yet been documented among the relatively few cases reported in multiple studies in the literature [5].
Problems/Controversies Unique to Mucinous BOTs
In addition to the problems discussed above which are unique to serous BOTs, a different group is associated with the mucinous intestinal type of BOT. These include (1) intraepithelial carcinoma, (2) the distinction from metastases to the ovaries, and (3) the problem of pseudomyxoma peritonei.
Intraepithelial Carcinoma
Although one might argue that the noninvasive micropapillary/cribriform growth pattern in a serous BOT (as described and discussed above) represents an intraepithelial carcinoma, in fact this diagnostic term has been used only in the setting of mucinous intestinal BOTs. The origin of this concept dates back more than 3 decades, when Hart and Norris [38] first stated that mucinous BOTs with more than 3 layers of cellular stratification should be classified as carcinomas. They also noted that such foci in mucinous intestinal BOTs were also characterized by marked nuclear atypia, and gradually this latter criterion has replaced a specific degree of stratification as the defining criterion for intraepithelial carcinoma [4,6].
A minority of investigators have also considered intraglandular cribriform or stroma-free papillary growth as a diagnostic criterion [6]. By definition, of course, these lesions lack invasion of the underlying stroma. It is not clear from the literature—and the BOT workshop did not address the question of—whether there is a threshold amount of severely atypical proliferation below which the diagnosis of intraepithelial carcinoma is not recommended.
Although the diagnosis of intraepithelial carcinoma is used by most experts and was recommended by the BOT workshop [3,4,6,14,20], it is my personal belief that the use of this term adds little to the diagnosis and management of mucinous BOTs and introduces the term “carcinoma,” which may lead to over treatment, in a situation in which this is not necessary. In our own study [39], we noted no difference in survival or recurrence in patients with mucinous intestinal BOTs (all patients with follow-up were alive without evidence of recurrence) regardless of whether they were classified as grade 1, 2, or 3 based on criteria of stratification and cytologic atypia. In an extensive review of the literature, we found that the published recurrence and/or death rate for women with stage IA mucinous intestinal BOTs was 3.1% [39]. A more recent review of cases of stage I mucinous BOTs with intraepithelial carcinoma noted 18 deaths due to disease among 290 cases (6.2%) [6]. Considering that some of these cases in each series probably represented either metastases from occult primary tumors elsewhere or primary ovarian tumors with foci of destructive invasion present in unsampled tissue, it is probable that the real rate of untoward events in true mucinous intestinal BOTs is even lower than reported in these reviews and that the difference between those with and without intraepithelial carcinoma is so minimal that it does not, in my own opinion, justify a change in terminology for those tumors with the latter appearance. It would seem prudent, however, to recommend that more extensive sampling be done in any case with focal severe atypia (or, for that matter, with microinvasion) to attempt to rule out occult invasive carcinoma.
Metastatic Mucinous Carcinoma to the Ovary
Although it is possible that a rare case of serous or seromucinous BOT may be mimicked by a metastasis to the ovary, for practical purposes it is only with the mucinous intestinal type of BOT that this differential diagnosis deserves serious consideration. It has been pointed out by numerous observers that metastatic mucinous adenocarcinomas in the ovary, particularly but not exclusively from the gastrointestinal tract, can mimic not only primary mucinous adenocarcinoma but also mucinous BOT and even primary benign mucinous ovarian tumors [4,6,10,20,40-42]. Different authors have recommended somewhat variable criteria for distinguishing these occult metastases, but bilaterality, small size (typically less than 10 cm), ovarian surface involvement, and a multinodular pattern of involvement are criteria which are almost universally accepted, while a “garland” or cribriform pattern and “dirty” necrosis are somewhat more controversial [6,40-42]. Obviously, the known presence of an extraovarian primary mucinous carcinoma is diagnostically useful, and immunohistochemistry is also of considerable value [43,44]. In primary mucinous intestinal ovarian tumors, cytokeratin (CK) 7 is virtually always positive, and CK 20 is positive in about half of the cases; on the other hand, in metastatic colorectal carcinomas, CK 20 is virtually always positive and CK 7 negative. Ji and colleagues [44] have also recommended the use of antibody Dpc4 in the identification of metastatic pancreatic adenocarcinomas. It is worth noting that, in a routine hospital-based (as opposed to consultation) practice, Seidman and colleagues [42] found that approximately three-fourths of mucinous adenocarcinomas in the ovaries were metastatic rather than primary. See the article by Vang and Ronnett in this issue for a thorough discussion of this problem.
Pseudomyxoma Peritonei
For many years, the clinical syndrome of pseudomyxoma peritonei (defined as the presence of mucinous ascites containing variable numbers of mucinous cells and glands) was thought to represent the usual manifestation of peritoneal dissemination of mucinous intestinal BOTs. However, numerous clinical, pathologic, immunohistochemical, and molecular studies over the past decade have established that this syndrome is rarely if ever associated with primary ovarian tumors and is almost always the result of a ruptured low-grade mucinous tumor of the appendix [3,4,6,14,41,45]. Indeed, the finding of mucinous ascites in a patient with an apparent mucinous intestinal BOT should be considered strong presumptive evidence that the ovarian tumor actually represents a metastasis. In view of this, the BOT workshop recommended that ovarian tumors in this situation “should not be labeled with the same diagnostic terms used for primary ovarian tumors (cystadenoma or borderline), to avoid confusing clinicians about the pathologic interpretation of the established or suspected primary site.” It was also recommended, “when intraoperative consultation with frozen section leads to the diagnosis of a mucinous ovarian tumor in the setting of PMP, the suggestion for appendectomy should be conveyed to the surgeon, and the pathologist should examine the entire appendix microscopically.” [3]
Although it is thus beyond the scope of this article to discuss pseudomyxoma peritonei in greater detail, it should be remembered that this is essentially a clinical term, and the pathologic diagnosis is best rendered (as suggested initially by Ronnett and colleagues [46] as “disseminated peritoneal adenomucinosis” when the mucinous epithelial component of the lesion is sparse and low grade or “peritoneal mucinous carcinomatosis” when histologically malignant mucinous epithelium is easily found in the mucinous ascitic material.
Seromucinous/Endocervical/Müllerian Mucinous BOT
These terms, although accepted by most of the participants in the BOT workshop as synonymous, nevertheless may actually imply some subtle differences. Since the original report of Rutgers and Scully [47], only a few additional series of cases have been reported [39,48-50], and thus our understanding of this group of tumors is even less than that of the serous and mucinous intestinal BOTs. The classic mucinous BOT of endocervical type presents grossly as a unilocular or paucilocular cystic tumor with scattered grossly visible intracystic papillae and microscopically as a papillary tumor with architectural features similar to those of the usual serous BOT but with a lining composed predominantly of mucinous cells with an endocervical rather than an intestinal appearance and, by definition, lacking goblet cells. A marked stromal neutrophilic infiltrate is also characteristic of these tumors. Unlike mucinous intestinal BOTs, these tumors are frequently (up to 50%) bilateral, are generally smaller than the typical mucinous intestinal BOT, may be associated with endometriosis, and behave like serous BOTs, with similar patterns of peritoneal “implants.” They may contain variable proportions of serous, endometrioid, and other epithelia, and it is unclear whether the presence of these mixed patterns leads to behavior different than that described for the classic mucinous müllerian or endocervical BOT. Both microinvasive [39,48-50] and fully invasive [48-51] variants of this tumor have been reported, and both Dube et al [48] and Rodriguez et al [49] have suggested criteria for intraepithelial carcinoma. To date, tumor-related death has never been reported in a borderline tumor of this variety.
Summary
This review has discussed the spectrum of serous and mucinous BOTs, with emphasis on current problems and controversies in this rapidly changing field. The reader is also referred to the review of Sherman and colleagues [7] for a discussion of limitations in existing research and opportunities for future advances.
REFERENCES
1. Taylor HC Jr. Malignant and semimalignant tumors of the ovary. Surg Gynecol Obstet. 1929;48:204–230.
2. Berman JJ. Borderline ovarian tumor workshop, Bethesda, Maryland, August 27–28, 2003. Hum Pathol. 2004;35:907–909.
3. Silverberg SG, Bell DA, Kurman RJ, et al. Borderline ovarian tumors: key points and workshop summary. Hum Pathol. 2004;35:910–917.
4. Seidman JD, Soslow RA, Vang R, et al. Borderline ovarian tumors: diverse contemporary viewpoints on terminology and diagnostic criteria with illustrative images. Hum Pathol. 2004;35:918–933.
5. Bell DA, Longacre TA, Prat J, et al. Serous borderline (low malignant potential, atypical proliferative) ovarian tumors: workshop perspectives. Hum Pathol. 2004;35:934–948.
6. Ronnett BM, Kajdacsy-Balla A, Gilks CB, et al. Mucinous borderline ovarian tumors: points of general agreement and persistent controversies regarding nomenclature, diagnostic criteria, and behavior. Hum Pathol. 2004;35:949–960.
7. Sherman ME, Berman JJ, Birrer MJ, et al. Current challenges and opportunities for research on borderline ovarian tumors. Hum Pathol. 2004;35:961–970.
8. Serov SF, Scully RE, Sobin LH. Histologic typing of ovarian tumors. In: World Health Organization, International Histological Classification and Staging of Tumors. Geneva, Switzerland: World Health Organization; 1973.
9. Lee KR, Tavassoli FA, Prat J, et al. Tumors of the ovary and peritoneum: surface epithelial-stromal tumours. In: Tavassoli FA, Devilee P, eds. Pathology and Genetics of Tumours of the Breast and Female Genital Organs, World Health Organization Classification of Tumors. Lyon: IARC Press; 2003:113–145.
10. Scully RE, Young RH, Clement PB. Tumors of the Ovary, Maldeveloped Gonads, Fallopian Tube, and Broad Ligament: Atlas of Tumor Pathology, Third Series, Fascicle 23. Washington, DC: Armed Forces Institute of Pathology; 1998:51–168.
11. Russell P. Ovarian epithelial tumors with atypical proliferation. In: Low D, Fox H, eds. Advances in Gynaecological Pathology. New York, NY: Churchill Livingstone; 1992:299–320.
12. Seidman JD, Kurman RJ. Ovarian serous borderline tumors: a critical review of the literature with emphasis on prognostic indicators. Hum Pathol. 2000;31:539–557.
13. Bell DA. Origins and molecular pathology of ovarian cancer. Mod Pathol. 2005;18:S19–S32.
14. Hart WR. Borderline epithelial tumors of the ovary. Mod Pathol. 2005;18:S33–S50.
15. Purdie DM, Webb PM, Siskind V, et al. The different etiologies of mucinous and nonmucinous ovarian cancers. Gynecol Oncol. 2002;88:S145–S148.
16. Riman T, Dickman PW, Nilsson S, et al. Risk factors for epithelial borderline ovarian tumors: results of a Swedish case-control study. Gynecol Oncol. 2001;83:575–585.
17. Hoerl MD, Hart WR. Primary ovarian mucinous cystadenocarcinomas: a clinicopathologic study of 49 cases with long-term follow-up. Am J Surg Pathol. 1998;22:1449–1462.
18. Nayar R, Siriaunkgul S, Robbins KM, et al. Microinvasion in low malignant potential tumors of the ovary. Hum Pathol. 1996;27:521–527.
19. Gilks CB. Subclassification of ovarian surface epithelial tumors based on correlation of histologic and molecular pathologic data. Int J Gynecol Pathol. 2004;23:200–205.
20. Hart WR. Mucinous tumors of the ovary: a review. Int J Gynecol Pathol. 2004;24:4–25.
21. Burks RT, Sherman ME, Kurman RJ. Micropapillary serous carcinoma of the ovary: a distinctive low-grade carcinoma related to serous borderline tumors. Am J Surg Pathol. 1996;20:1319–1330.
22. Smith-Sehdev AE, Sehdev PS, Kurman RJ. Noninvasive and invasive micropapillary (low-grade) serous carcinoma of the ovary: a clinicopathologic analysis of 135 cases. Am J Surg Pathol. 2003;27:725–736.
23. Segal GH, Hart WR. Ovarian serous tumors of low malignant potential (serous borderline tumors): the relationship of exophytic surface tumor to peritoneal “implants”. Am J Surg Pathol. 1992;16:577–583.
24. Staebler A, Heselmeyer-Haddadd K, Bell K, et al. Micropapillary serous carcinoma of the ovary has distinct patterns of chromosomal imbalances by comparative genomic hybridization compared with atypical proliferative serous tumors and serous carcinomas. Hum Pathol. 2002;33:47–59.
25. Singer G, Shih I-M, Truskinovsky A, et al. Mutational analysis of K-ras segregates ovarian serous carcinoma into two types: invasive MPSC (low-grade tumor) and conventional serous carcinoma (high grade tumor). Int J Gynecol Pathol. 2002;22:37–41.
26. Gu J, Roth LM, Younger C, et al. Molecular evidence for the independent origin of extra-ovarian papillary serous tumors of low malignant potential. J Natl Cancer Inst. 2001;93:1147–1152.
27. Lu KH, Bell DA, Welch WR, et al. Evidence for the multifocal origin of bilateral and advanced human serous borderline ovarian tumors. Cancer Res. 1998;58:2328–2330.
28. Ortiz BH, Ailawadi M, Colitti C, et al. Second primary or recurrence? comparative patterns of p53 and K-ras mutations suggest that serous borderline ovarian tumors and subsequent serous carcinomas are unrelated tumors. Cancer Res. 2001;61:7264–7267.
29. Sieben NLG, Kolkman-Uljee SM, Flanagan AM, et al. Molecular genetic evidence for monoclonal origin of bilateral ovarian serous borderline tumors. Am J Pathol. 2003;162:1095–1101.
30. Bostwick DG, Tazelaar HD, Ballon SC, et al. Ovarian epithelial tumors of borderline malignancy: a clinical and pathologic study of 109 cases. Cancer. 1986;58:2052–2065.
31. Michael H, Roth LM. Invasive and noninvasive implants in ovarian serous tumors of low malignant potential. Cancer. 1986;57:1240–1247.
32. Gershenson DM, Silva EG. Serous ovarian tumors of low malignant potential with peritoneal implants. Cancer. 1990;65:578–585.
33. Bell DA, Weinstock MA, Scully RE. Peritoneal implants of ovarian serous borderline tumors: histologic features and prognosis. Cancer. 1988;62:2212–2222.
34. Bell KA, Smith-Sehdev AE, Kurman RJ. Refined diagnostic criteria for implants associated with ovarian atypical proliferative serous tumors (borderline) and micropapillary serous carcinomas. Am J Surg Pathol. 2001;25:419–432.
35. Silva EG, Tornos C, Zhuang Z, et al. Tumor recurrence in stage I ovarian serous neoplasms of low malignant potential. Int J Gynecol Pathol. 1998;17:1–6.
36. Moore WF, Bentley RC, Berchuck A, Robboy SJ. Some müllerian inclusion cysts in lymph nodes may sometimes be metastases from serous borderline tumors of the ovary. Am J Surg Pathol. 2000;24:710–718.
37. Alvarez AA, Moore WF, Robboy SJ, et al. K-ras mutations in müllerian inclusion cysts associated with serous borderline tumors of the ovary. Gynecol Oncol. 2001;80:201–206.
38. Hart WR, Norris HJ. Borderline and malignant mucinous tumors of the ovary: histologic criteria and clinical behavior. Cancer. 1973;31:1031–1045.
39. Siriaunkgul S, Robbins KM, McGowan L, et al. Ovarian mucinous tumors of low malignant potential: a clinicopathologic study of 54 tumors of intestinal and müllerian type. Int J Gynecol Pathol. 1995;14:198–208.
40. Lee KR, Young RH. The distinction between primary and metastatic mucinous carcinomas of the ovary: gross and histologic findings in 50 cases. Am J Surg Pathol. 2003;27:281–292.
41. Riopel MA, Ronnett BM, Kurman RJ. Evaluation of diagnostic criteria and behavior of ovarian intestinal-type mucinous tumors: atypical proliferative (borderline) tumors and intraepithelial, microinvasive, invasive, and metastatic carcinomas. Am J Surg Pathol. 1999;23:617–635.
42. Seidman JD, Kurman RJ, Ronnett BM. Primary and metastatic mucinous adenocarcinomas in the ovaries: incidence in routine practice with a new approach to improve intraoperative diagnosis. Am J Surg Pathol. 2003;27:985–993.
43. Cathro HP, Stoler MH. Expression of cytokeratins 7 and 20 in ovarian neoplasia. Am J Clin Pathol. 2002;117:944–951.
44. Ji H, Isacson C, Seidman JD, et al. Cytokeratins 7 and 20, Dpc4, and MUC5AC in the distinction of metastatic mucinous carcinomas in the ovary from primary ovarian mucinous tumors: Dpc4 assists in identifying metastatic pancreatic carcinomas. Int J Gynecol Pathol. 2002;21:391–400.
45. Ronnett BM, Kurman RJ, Shmookler BM, et al. The morphologic spectrum of ovarian metastases of appendiceal adenocarcinomas: a clinicopathologic and immunohistochemical analysis of tumors often misinterpreted as primary ovarian tumors or metastatic tumors from other gastrointestinal sites. Am J Surg Pathol. 1997;21:1144–1155.
46. Ronnett BM, Zahn CM, Kurman RJ, et al. Disseminated peritoneal adenomucinosis and peritoneal mucinous carcinomatosis: a clinicopathologic analysis of 109 cases with emphasis on distinguishing pathologic features, site of origin, prognosis, and relationship to “pseudomyxoma peritonei”. Am J Surg Pathol. 1995;19:1390–1408.
47. Rutgers J, Scully RE. Ovarian müllerian mucinous papillary cystadenomas of borderline malignancy: a clinicopathologic analysis. Cancer. 1998;61:340–348.
48. Dube V, Roy M, Plante M, et al. Mucinous ovarian tumors of müllerian-type: an analysis of 17 cases including borderline tumors and intraepithelial, microinvasive, and invasive carcinomas. Int J Gynecol Pathol. 2005;24:138–146.
49. Rodriguez IM, Irving JA, Prat J. Endocervical-like mucinous borderline tumors of the ovary: a clinicopathologic analysis of 31 cases. Am J Surg Pathol. 2004;28:1311–1318.
50. Shappell HW, Riopel MA, Smith-Sehdev AE, et al. Diagnostic criteria and behavior of ovarian seromucinous (endocervical-type mucinous and mixed cell-type) tumors: atypical proliferative (borderline) tumors, intraepithelial, microinvasive, and invasive carcinomas. Am J Surg Pathol. 2002;26:1529–1541.
51. Lee KR, Nucci MR. Ovarian mucinous and mixed epithelial carcinomas of müllerian (endocervical-like) type: a clinicopathologic analysis of four cases of an uncommon variant associated with endometriosis. Int J Gynecol Pathol. 2003;22:42–51.
Pathology Case Reviews, Volume 11(1), January/February 2006, pp 9-17
BOTTOM LINE
Museum In India Tackles HIV/AIDS
Manjulika Das
A municipality in western India is using a sex museum to educate people about HIV/AIDS and safe sex.
The Antarang museum in Mumbai uses excerpts from India's ancient books, including the Kama Sutra and Kumarsambhabam, to describe the sexual relationship between men and women. Paintings and sculpture depict the basics of love, reproduction and HIV/AIDS. Exhibits also illustrate condom use and safe sex, and discuss the dangers of unprotected sex, the prevalence of HIV/AIDS in India and its social stigma.
An estimated 5.13 million people in India are infected with HIV, the highest number of infections in the world after South Africa (CMAJ 2004; 171:1337-8). The southern states, where Mumbai is located, account for 30% of India's population but 75% of HIV-1 cases.
The museum is a collaborative venture by the Brihanmumbai Municipal Corporation and the Mumbai District AIDS Control Society. When it opened in 2002, it attracted some 25 000 visitors annually. Now, the numbers have tapered off, largely due to a lack of publicity and the social stigma surrounding HIV/AID in India. Precise attendance figures are not available.
Nonetheless, the museum is effective, maintains Swapan Jana, secretary of society for Social Pharmacology India, a non-governmental organization working on HIV/AIDS education and prevention. "Sex education provided by the museum has really helped save lives of people," he says.
If Antarang's operators maintain the museum and pitch it toward young people, it may be more successful than traditional approaches to combating HIV/ AIDS, he adds.
"In India, approaches carried out with posters, leaflets, free condoms, auto-disposable syringes and television publicity seem to have very limited impact in combating HIV/AIDS, because most of them couldn't attract the common people," says Jana. "Antarang is an attractive approach [and] it should work properly, reaching the target population." — Manjulika Das, Kolkata, India
Canadian Medical Association Journal, January 3, 2006; 174 (1).