Mucinous Tumors of the Ovary: A Clinicopathologic Study of 196 Borderline Tumors (of Intestinal Type) and Carcinomas, Including an Evaluation of 11 Cases With `Pseudomyxoma Peritonei'

Lee, Kenneth R. M.D.; Scully, Robert E. M.D.

American Journal of Surgical Pathology:
Original Articles

Mucinous ovarian neoplasms other than cystadenomas and adenofibromas have been classified as either borderline tumors or carcinomas for many years. Borderline tumors have been subdivided more recently into endocervical-like (mullerian) and intestinal forms. Such a distinction is rarely made in the mucinous carcinoma category. We did not encounter a pure endocervical-like carcinoma in the present series. Criteria for distinguishing an intestinal-type mucinous borderline tumor from a mucinous carcinoma have been controversial. In this study of 164 mucinous borderline tumors of intestinal type and 32 mucinous carcinomas, the former were further subdivided into 74 cases with epithelial atypia only and 90 with focal intraepithelial carcinoma. Of the 67 stage I tumors in the borderline (with atypia) category, all 49 with follow-up data were clinically benign; in the seven cases that had been designated stage III, the intraoperative appearance was that of [ldquo]pseudomyxoma peritonei,[rdquo] which was fatal in four cases. Most of these tumors, however, were probably metastatic to the ovary rather than truly primary borderline tumors, although failure to examine the appendix in six cases compromised their interpretation. All 90 mucinous borderline tumors that had foci of intraepithelial carcinoma were recorded as stage I, but two of the 69 patients with follow-up data (3[percnt]) had fatal recurrences. Both of these tumors were incompletely staged, however, and one had ruptured intraoperatively. Thirty-two invasive carcinomas were subdivided into 12 expansile and 20 infiltrative subtypes; within the latter category seven tumors were only microinvasive. All 12 carcinomas with only expansile invasion were stage I; none of the 10 with follow-up data recurred. All seven microinvasive infiltrative carcinomas were stage I; none of the five with follow-up data recurred. One of five patients with stage I infiltrative carcinomas that were more than microinvasive and were adequately followed had a fatal recurrence, but staging had been incomplete in that case. Seven of the remaining eight infiltrative carcinomas were higher than stage I; five of the six (83[percnt]) with follow-up data persisted or recurred and were fatal. Considering all stages, increasing tumor grade in the carcinoma category correlated with an unfavorable outcome. However, grade did not influence prognosis in stage I carcinomas. Among 13 stage I cases in all categories with either preoperative or intraoperative tumor rupture and follow-up data, one recurred, a tumor in the borderline with intraepithelial carcinoma category. [ldquo]Pseudomyxoma peritonei[rdquo] is an ill-defined term and should not be used as a pathologic diagnosis. The presence of mucin in the abdominal cavity requires careful histologic evaluation to characterize it for prognostic purposes. Adequate and sometimes extensive sampling of mucinous ovarian tumors, the appendix and the peritoneum in cases of [ldquo]pseudomyxoma peritonei[rdquo] is necessary to achieve an accurate diagnosis and prognosis.

Author Information

From the Department of Pathology, Brigham and Women's Hospital, and the James Homer Wright Pathology Laboratories of the Massachusetts General Hospital, Boston, Massachusetts, U.S.A.

Address correspondence and reprint requests to Kenneth R. Lee, MD, Department of Pathology, Brigham and Women's Hospital, 75 Francis St., Boston, MA 02115, U.S.A.; e-mail[colon] KRLEE[commat]BICS.BWH.HARVARD.EDU

Article Outline

Thirty percent to 50[percnt] of ovarian epithelial borderline tumors and 6[percnt] to 10[percnt] of ovarian carcinomas are of the mucinous type. 51 The major criterion proposed by the World Health Organization (WHO) in 1973 and reiterated in 1999 for separation of mucinous borderline tumors and mucinous carcinomas (obvious invasion of the stroma in the latter) 53,54 has not been accepted by all investigators, 16,17 making comparison of outcomes in reported series of cases difficult. Also, since the publication of the original WHO [ldquo]Histological Typing of Ovarian Tumors,[rdquo] several studies of ovarian mucinous tumors have contributed new insights and introduced additional controversies concerning these neoplasms[colon] (1) Mucinous borderline tumors have been separated into intestinal and endocervical-like (mullerian) subtypes, which differ in both their histologic features and their clinical behavior. 46,56 (2) Grading of the epithelium in intestinal-type borderline tumors, primarily based on the degree of nuclear atypia, has been studied as a possible prognostic factor. 15,44,56,57,63 (3) On the basis of a small number of cases, evidence has been presented that microinvasion of the stroma in otherwise borderline mucinous tumors is not clinically significant. 15,18,22,30,39 (4) Stage I mucinous carcinomas have been shown to have a better prognosis than previously thought, 15,18,22,39,43,48,63 possibly because metastatic tumors to the ovary, particularly from the intestine and pancreas, may have been misinterpreted as primary mucinous carcinomas in early reported series. 39,40,51 (5) Evidence has also been proffered that all or virtually all ovarian mucinous cystic tumors associated with [ldquo]pseudomyxoma peritonei[rdquo] (PP) and considered primary, according to the International Federation of Gynecology and Obstetrics (FIGO) staging system, are actually metastatic from the appendix or another gastrointestinal site 41,54,65; in other words, PP in women is never, or almost never, a complication of a primary mucinous tumor of the ovary. 41[ndash]43

In an attempt to contribute data germane to these new insights and controversies and to adhere to current WHO and FIGO principles of classification and staging, we have conducted a clinicopathologic study of 196 cases of mucinous borderline tumors of intestinal type and mucinous carcinomas of the ovary. On the basis of our findings, we propose a subclassification of these tumors consistent with and expansive of the WHO classification, discuss their current FIGO staging and its shortcomings, and propose new staging criteria for consideration by staging committees.

Back to Top | Article Outline


All cases coded as mucinous carcinoma or mucinous borderline tumor in the consultation files of one of the authors (R.E.S.; 1967[ndash]1985) and in the surgical pathology and tumor registry files of the Brigham and Women's Hospital (1980[ndash]1991) were retrieved and analyzed. Endocervical-like mucinous borderline neoplasms, 46 which accounted for 12[percnt] of the borderline tumors in the hospital series, and mucinous tumors containing nodules of anaplastic carcinoma or sarcoma, 33,34 several of which were encountered in the consultation series, were excluded from the investigation. Carcinomas considered most likely metastatic based on clinical or pathologic features, or both, were also excluded. 39,40,51 Cases of [ldquo]ovarian borderline tumors[rdquo] in which a clinical or surgical diagnosis of PP was made were retained in the series, however, in accordance with the current FIGO practice of designating such tumors primary. One hundred ninety-six cases, 131 from the consultation series and 65 from the hospital series, remained for analysis. All available slides were reviewed. The number of sections per tumor ranged from 1 to 36 (mean, 10; median, 8) in the consultation series and from 4 to 46 (mean, 18; median, 17) in the hospital series. The mean number of sections per centimeter of tumor diameter available for review was 0.6 in the hospital series and 0.4 in the consultation series. The latter figure is probably lower than that for the original sampling of the tumors because in an unknown number of cases only the slide or slides showing areas that were suspicious for the highest degree of malignancy were submitted for consultation and available for review. Information regarding the clinical presentation, tumor size and appearance, stage, treatment, and follow up was obtained from the clinical records or by oral or written communication with pathologists, clinicians, and tumor registries.

The tumors were divided into three major histologic categories and several subcategories. Noninvasive tumors with intraglandular or intracystic epithelial proliferation in the absence of disturbing architectural features of the lining epithelium, as listed below, and with only slight (grade 1) nuclear atypia, were designated simply [ldquo]borderline[rdquo] (borderline of Hart and Norris) 16 (Figs. 1 and 2). If a borderline tumor contained foci in which the epithelial lining cells had moderately (grade 2) or highly atypical (grade 3) nuclei, or were stratified into four or more layers or had a cribriform or stroma-free papillary growth pattern, or a combination of these features, the designation [ldquo]borderline with intraepithelial carcinoma[rdquo] was applied (Figs. 3, 4, and 5). Neoplasms with stromal invasion were classified as carcinomas. Two forms of stromal invasion were recognized. One, designated the infiltrative type, was characterized by disorderly penetration of the stroma by neoplastic glands, single cells, or cell clusters, sometimes eliciting a desmoplastic stromal response (Fig. 6). This type of invasion ranged quantitatively from minor to predominant within the specimen. The term [ldquo]microinvasive carcinoma[rdquo] was assigned to cases in which the invasion was limited to one or more foci of infiltrative invasion 10 square mm or less in area. The second invasive pattern, designated the expansile type, was characterized by a complex, often labyrinthine, arrangement of glands, cysts, or papillae lined by malignant epithelium with minimal or no intervening stroma recognizable on light microscopic examination; the expansile pattern of growth had to exceed 10 square mm in area and be at least 3 mm in each of two linear dimensions to qualify as invasive carcinoma (Figs. 7 and 8). Carcinomas with expansile or infiltrative invasive foci composed entirely of glands, cysts, or cysts with papillae were designated grade 1; those containing up to 50[percnt] solid foci, grade 2; and those containing 50[percnt] or more solid foci, grade 3. If severe nuclear atypia was present in a carcinoma that was otherwise grade 1 or 2, the grade was raised by one. Tumors with only microinvasion were not graded.

Within the limits imposed by the variable number of slides per case, several histologic features were examined in detail. The percentages of each tumor that were composed of cystadenoma, borderline (with atypia), borderline with intraepithelial carcinoma, and invasive carcinoma were estimated. The following findings were quantified as rare, few, or many according to whether there were two or fewer foci, three to five foci, or more than five foci, respectively[colon] nuclear stratification greater than three layers; a cribriform pattern (Figs. 3 and 4); and in intraepithelial carcinomas, cells with grade 3 nuclear features (Fig. 5). Mitotic figures were counted in at least 10 high-power (400[times]) fields (HPF) in the most mitotically active areas and quantitated as follows[colon] [le]4/10 HPF[mdash]few; 5[ndash]9/10 HPF[mdash]moderate numbers; [ge]10/10 HPF[mdash]many. Pseudomyxoma ovarii (PO) was defined as more than a minor focus of mucin dissecting into the stroma with or without epithelial cells and without an accompanying histiocytic or other inflammatory cell reaction (Fig. 9).

The term [ldquo]pseudomyxoma peritonei[rdquo] (PP) was defined clinically on the basis of intraoperative findings as a grossly visible, localized, or generalized accumulation of mucus in the pelvis or abdomen, either lying on and attached to the peritoneal surfaces or incorporated within dense fibrous tissue. On microscopic examination, the intraabdominal mucin was subdivided into three categories[colon] (1) superficial organizing type[mdash]adherent mucin confined to the peritoneal surfaces, containing organizing capillaries, fibroblasts, and one or more of the following additional cells[colon] epithelial, mesothelial, or inflammatory (Fig. 10); (2) dissecting with fibrosis type[mdash]pools of mucin with or without epithelial cells surrounded by dense collagenous tissue (Figs. 11 and 12); or (3) metastatic mucinous carcinoma (with or without mucin dissection and fibrosis)[mdash]mucin associated with mucin-producing carcinoma. In each of the first two categories, the presence or absence of epithelial cells on microscopic examination and, if detected, whether they had benign or borderline (atypical) features were recorded.

Tumor stage was determined retrospectively by the authors according to the FIGO system. 32 Cases in which the extent of the initial surgical exploration was not specified but no reference to extra-ovarian spread was made in the consultation letter or pathology report were considered stage I. Tumors accompanied by PP in which the mucin was superficial (without dissection into tissues on microscopic examination) were considered stage Ic. Those with dissecting mucin with or without cells were considered stage II or III. In two cases in which the intraabdominal mucus was not sampled, the stage was determined based on the operative characterization of the mucus. If it was described as forming intraabdominal masses within tissue or solid surface deposits, the tumor was considered stage III.

For all categories combined, 5-year survival rates relative to stage were computed according to the log rank test. Within each category and subcategory of carcinoma, the clinical outcome was correlated with tumor stage, the presence of PO, and the occurrence of tumor rupture. In cases of stage I borderline tumors with intraepithelial carcinoma, an attempt was made to correlate the outcome with the percentage of the tumor containing intraepithelial carcinoma and with the percentage of grade 3 nuclei. Seven stage I patients were known to have received postoperative chemotherapy. However, because data regarding possible postoperative radiation or chemotherapy in other patients are unavailable, adjuvant therapy was not included in the outcome analysis.

Back to Top | Article Outline


Of the 196 cases, 74 (38[percnt]) were borderline (with atypia), 90 (46[percnt]) were borderline with intraepithelial carcinoma, 7 (4[percnt]) were borderline with intraepithelial carcinoma and one or more foci of microinvasion, and 25 (12[percnt]) were invasive carcinomas (13 infiltrative with or without additional expansile invasion and 12 purely expansile). On gross examination, most of the tumors were described as smooth-surfaced, multiloculated cystic masses. In six of the 11 cases with PP, surface [ldquo]fibrinous[rdquo] or mucoid excrescences were described. Similar descriptions were recorded in five of the 185 cases without PP. Solid foci were noted in 35[percnt] and papillary foci in 10[percnt] of the cases. Thirty-two tumors (16[percnt]) were unilocular cysts, with most of them containing at least one solid or predominantly solid focus; all four unilocular cysts that proved on microscopic examination to be carcinomas had a predominantly solid nodule described as multicystic, gelatinous, or papillary in their walls. The age range in the consultation series was 14 to 84 years (average, 34 yrs) and in the hospital series, 18 to 81 years (average, 44 yrs). Seven tumors (4[percnt]) were bilateral, all in the borderline-with-atypia category and five of these were associated with PP; 84 were left-sided and 83 were right-sided; in 22 unilateral cases the side was not specified. The median largest linear dimension of the tumors was 19 cm (range, 6[ndash]48 cm).

Seven tumor specimens had a component in another diagnostic category; in five of them (2.5[percnt]) it was a dermoid cyst and in two (1[percnt]) it was a Brenner tumor. Sixteen tumors (8.2[percnt]) were accompanied by a benign neoplasm in the contralateral ovary[colon] 4 mucinous cystadenomas, 7 serous adenofibromas, 3 dermoid cysts (including one case in which the mucinous tumor in the ipsilateral ovary was mixed with a dermoid cyst), and 2 Brenner tumors. The contralateral ovary contained endometriotic foci in three cases. One hundred eighty-two tumors (93[percnt]) were stage I; 16 of these were stage Ic because of positive peritoneal washings, surface involvement, rupture, PP of the superficial organizing type, or various combinations of these findings. Three tumors were stage II, 10 were stage III, and 1 was stage IV. Patient age and staging data are summarized in Table 1.

Follow-up information was obtained in 151 of the 196 cases (77[percnt];Table 2). The mean follow-up interval in the patients without recurrence was 8 years (range, 2[ndash]26 yrs). Tumor persisted or recurred in 12 of the 151 patients (8[percnt]); and 11 of them died of their disease; one patient with persistent disease died of other causes. Sixteen patients died of other causes without clinical evidence of tumor at the time of their death.

Back to Top | Article Outline
Borderline Tumors (With Atypia; 74 Cases)

The mean age of the patients was 46 years in both the consultation and the hospital series (range, 14[ndash]84 yrs). Of the 74 tumors, 59 were stage Ia, 2 stage Ib, 6 stage Ic, and 7 stage III. In six cases the tumors had ruptured preoperatively or intraoperatively; in two of these cases there was associated PP. The atypia involved more than half of the epithelial area in most of the cases, with the remaining areas having the features of a cystadenoma; 16[percnt] of the tumors were entirely atypical, and in two cases the atypical areas accounted for less than 10[percnt] of the epithelial area (Figs. 1 and 2). Foci with many mitotic figures were seen in 13[percnt] of the cases, 37[percnt] of the tumors had moderate numbers of mitotic figures, and 50[percnt] had only a few. Pseudomyxoma ovarii was present in 23 cases (31[percnt]), including all 9 associated with PP (100[percnt]) and 14 of the 65 cases without PP (22[percnt]); in eight of the tumors with PO (35[percnt]), the mucinous pools contained tumor cells (Fig. 9). In six tumors without PO, there were small foci of stromal mucin with a histiocytic reaction, presumably caused by rupture of a mucin filled cyst. In nine tumors, there were large areas of necrosis and acute inflammation. A pronounced filiform pattern of growth, characterized by slender intracystic papillae with narrow stromal cores, was seen in 3 cases, stromal luteinization in 7, prominent Paneth cells in 1, and stromal calcification in 2 cases.

Follow-up information was obtained in 56 cases (41 stage Ia, 2 stage Ib, 6 stage Ic, and 7 stage III). In 18 of the 49 stage I cases, the surgical treatment was unilateral oophorectomy (17 cases) or cystectomy (1 case) with conservation of the contralateral ovary. No stage I tumors recurred (0 of 49); two of them were stage Ic because the intraoperative impression (massive mucinous or gelatinous ascites with multiple mucoid peritoneal [ldquo]implants[rdquo] in both) was microscopically determined to be of the superficial organizing type of PP (Tables 2 and 3); the specimens containing both of these tumors had an additional component of dermoid cyst.

All seven patients with stage III tumors had intraoperative evidence of PP (Tables 2 and 3). Four of these patients died of, or with, persistent PP in periods ranging from 9 months to 6 years 3 months. In all four fatal cases the mucinous tumors were bilateral; three of them had ovarian surface involvement; all of them contained only slightly atypical tall mucinous cells with similar cells in foci of PO (Fig. 12). In one of the four cases, a synchronous mucinous tumor of the appendix was removed; it had a histologic appearance similar to that of the ovarian tumor and had dissected into the appendiceal wall; in the other three cases the appendix was not removed or mentioned in the operative report. In at least three of the four cases, the PP was of the dissecting with fibrosis type and contained mucinous cells similar to those seen in the ovarian tumors. Slides of the PP were not available in the fourth case; the type of PP was not discernible from the original pathology report but it stated that tumor cells were not found in the intraperitoneal mucin. Of the three surviving patients with stage III tumors, one had bilateral ovarian tumors with intraoperative evidence of PP, which was not sampled for microscopic examination; both tumors had PO containing epithelial cells but surface involvement was not described grossly or seen microscopically; the appendix was not removed or mentioned in the operative report; the patient was alive without clinical evidence of recurrence at 4 years 10 months and subsequently was lost to follow up. The second surviving patient had a 25-cm unilateral left-sided tumor which had PO containing epithelial cells; the omentum, which was removed, contained numerous small, firm mucoid areas that were of the dissecting-with-fibrosis type histologically but no epithelial cells were seen in the sections examined; the appendix was not removed or mentioned in the operative report; however, a preoperative barium enema was reported as normal; the patient was alive without clinical evidence of recurrence after 8 years and was subsequently lost to follow up. The third surviving patient had a unilateral right-sided tumor with surface [ldquo]excrescences[rdquo] associated with a dermoid cyst and foci of PO without epithelial cells; there was preoperative rupture. The abdomen contained 6 L of gelatinous fluid, studding of all peritoneal surfaces with mucinous [ldquo]implants,[rdquo] and a soft fleshy tumor of the omentum. None of the abdominal mucoid material was sampled; the appendix was not removed or mentioned in the operative report. The patient was alive without clinical evidence of recurrence at last follow up 4 years 9 months after the initial operation.

Back to Top | Article Outline
Borderline Tumors With Intraepithelial Carcinoma (90 Cases)

The mean age of the patients was 36 years (43 in the hospital series and 34 in the consultation series; range, 18[ndash]81 yrs). Tumors in this category accounted for 52[percnt] of those in the consultation series and 37[percnt] of those in the hospital series. Eighty-nine of the tumors were stage Ia and one was stage Ic because of positive peritoneal washings. Six tumors were reported to be ruptured preoperatively or intraoperatively. The intraepithelial carcinomatous component ranged from less than 5[percnt] of the tumor in 37 (41[percnt]) of the cases to 90[percnt] in one case. Borderline (with atypia) foci were present in all the cases and cystadenoma in 70[percnt] of them, with the latter usually occupying less than 10[percnt] of the specimen. Seventeen tumors (19[percnt]) had foci of grade 3 nuclei (Fig. 5), cribriform foci were seen in 55[percnt] of the cases, and 50[percnt] had areas with many mitotic figures. PO was present in eight tumors; in one of these there were a few malignant-appearing epithelial cells in the dissecting mucin adjacent to a ruptured cyst containing similar cells. Mucin extravasation with a histiocytic response was present in 12 cases, extensive necrosis with acute inflammation in 13 cases, stromal luteinization in 4, prominent Paneth cells in 2, and focal stromal calcification in 3 cases.

Follow-up information was obtained in 69 cases. In 22 of the 69 cases, the surgical treatment was unilateral oophorectomy (21 cases) or cystectomy (1 case) with conservation of the contralateral ovary. There were two recurrences, both of which proved fatal (Tables 2 and 4). In the first case, an 18-year-old woman presented with a 30-cm left-sided tumor that required intraoperative drainage and ruptured during removal. She underwent unilateral oophorectomy and had a single peritoneal biopsy, which was negative for malignant cells. Inspection of the pelvis and abdomen revealed no other areas suspicious for malignancy. Twenty sections were taken (0.7 per cm tumor diameter); 20[percnt] of the specimen contained only cystadenoma, and almost all of the remainder was borderline tumor (with atypia). Several microscopic foci of intraepithelial carcinoma occupied less than 1[percnt] of the tumor and contained rare grade 3 nuclei. There were also rare foci of PO without epithelial cells and several areas of necrosis with mucin extravasation and acute inflammation. A computed tomography scan of the abdomen and pelvis 2 months postoperatively disclosed no evidence of tumor; the liver, spleen, pancreas, kidneys, uterus, and right ovary were described as normal. The patient presented 1 year later with obstructive jaundice and an enlarging cystic right pelvic mass. At exploration there were widespread metastases of invasive grade 2 mucinous adenocarcinoma without PP involving the abdominal wall, the serosa and muscularis of the descending colon, the uterine serosa, and the right ovary. Despite removal of these structures and a normal gallbladder, the patient died after 3 months. In the second case, a 54-year-old woman with a remote history of an appendectomy presented with an 18-cm left-sided ovarian tumor. She underwent a total abdominal hysterectomy with bilateral salpingo-oophorectomy, a punch biopsy of the ileal serosa, and a 4-cm biopsy of the omentum. Peritoneal washings contained neoplastic cells with grade 2 malignant nuclei but there was no intraoperative gross or histologic evidence of extra-ovarian spread. In the 20 sections examined (1.1 per cm tumor diameter), the ovarian tumor was predominantly borderline (with atypia), with an approximately 5[percnt] component of intraepithelial carcinoma containing rare grade 3 nuclei. There was extensive necrosis and acute inflammation in areas of mucin extravasation. Three years later the patient presented with bowel obstruction and an abdominal mass and underwent resection of the distal ileum, cecum, and right colon. Examination of the specimen showed multiple implants of grade 2 mucinous adenocarcinoma on the intestinal serosa and muscularis; no mucosal tumor was present. Despite subsequent chemotherapy, she died 2 months later with residual metastases involving the pelvis, abdomen, and liver.

In addition to these two cases, there were 12 other tumors containing grade 3 nuclei with follow-up data. None of them recurred. Because there were no other recurrences in this category, and because the two recurrent tumors were predominantly borderline with only minor foci of intraepithelial carcinoma, a difference in outcome based on the percentage of tumor that contained grade 3 nuclei or intraepithelial carcinoma was not demonstrable.

Back to Top | Article Outline
Borderline Tumors With Intraepithelial Carcinoma and Microinvasion (7 Cases)

The mean age of the patients was 37 years in the hospital series and 26 years in the consultation series (range, 24[ndash]66 yrs). Three of the 7 tumors had only 1 focus of invasion, 2 had 2, 1 had 10, and 1 had 15. In 3 of the 7 cases the component of intraepithelial carcinoma accounted for 5[percnt] or less of the tumor, with the remainder of the specimen being both borderline and benign in all three. All seven tumors were stage I. There were no recurrences in the five cases with follow-up data. In three of these cases, the surgical treatment was unilateral oophorectomy.

Back to Top | Article Outline
Mucinous Carcinomas (Invasive; 25 Cases)

The mean age of the patients was 44 years (50 in the 10 hospital cases, and 40 in the 15 consultation cases; range, 15[ndash]74 yrs). Three tumors, all stage I, were reported to have ruptured. In 12 (48[percnt]) of the 25 tumors, all stage Ia, the invasion was exclusively expansile. In 11 of these cases, benign, borderline, and intraepithelial carcinomatous areas accounted for 10[percnt] to 70[percnt] of the tumor; in the twelfth case, the entire tumor was expansively invasive. In all 12 of these tumors, major portions of the expansile malignant areas were similar to each other, characterized by a mixture of cystic, cribriform, and papillary glands lined by cells with grade 2 or 3 malignant nuclei and small to moderate amounts of nonspecific-appearing cytoplasm. Eight of these cases were classified as grade 1 and 4 as grade 2. In three tumors the expansile areas had foci of intestinal differentiation, and in two others there were endocervical-like areas. In the 10 cases with follow-up data, in three of which the surgical treatment was unilateral oophorectomy, there were no recurrences.

The remaining 13 carcinomas had either infiltrative invasion that was more than microinvasive (11 cases, 5 with additional foci of expansile invasion) or expansile invasion with foci of infiltrative microinvasion (2 cases). Six (46[percnt]) tumors were stage Ia, 2 stage IIa, 1 stage IIc, 3 stage III, and 1 stage IV. In contrast to the relative histologic uniformity of the malignant areas in the carcinomas with expansile invasion, the patterns of infiltrative invasion were diverse.

Three (23[percnt]) of the infiltrative tumors were pure carcinomas; one with a glandular and papillary pattern, and two in which the malignant cells invaded in the form of closely packed glands, cords, and clusters focally surrounded by extracellular mucin.

Nine (69[percnt]) of the infiltrative carcinomas contained foci of both borderline tumor with atypia and borderline tumor with intraepithelial carcinoma; 4 of the 9 also had benign-appearing foci. These nine tumors invaded in glandular, solid, or single cell arrangements with a wide range of nuclear atypicality and usually with a desmoplastic stromal response.

The thirteenth tumor was a grade 2 adenocarcinoma of intestinal type infiltrating as large and small glands in a desmoplastic stroma without benign or borderline mucinous foci and was associated with a dermoid cyst.

Three infiltrative carcinomas were grade 1, 5 grade 2, and 5 grade 3. There was no relationship between the grade and the stage within this carcinoma subgroup (data not shown). Follow-up data were obtained in 11 of the 13 cases. Four of the five patients with stage I tumors and follow-up data, two of whom were treated by unilateral oophorectomy, were free of recurrence in follow-up periods from 5 to 12 years. One patient with a stage I tumor died, but the staging had been limited to unilateral oophorectomy, several peritoneal biopsies, an omental biopsy, and a contralateral ovarian wedge biopsy, which showed endometriosis. A prior laparoscopic biopsy of the tumor, diagnosed as a mucinous borderline tumor, had preceded its removal. The 9-cm tumor was a mixture of borderline tumor with intraepithelial carcinoma, grade 1 expansile invasion, and two foci of infiltrative microinvasion, the larger 6 mm3 (4 [times] 1.5 mm) in area. Three years later the patient presented with ascites. Exploration revealed multiple implants of grade 3 mucinous carcinoma extensively involving the peritoneum, the entire uterus, the remaining ovary, the omentum, and the muscularis of the sigmoid colon. The patient died shortly after a posterior pelvic exenteration and institution of chemotherapy.

One patient with a stage IIa tumor survived disease-free for 11 years at last follow up. The other five patients with follow-up data whose tumors were higher than stage I (1 stage IIc, 3 stage III, 1 stage IV) all died of carcinoma in periods of up to 1 year (all with abdominal implants, 1 with liver metastases and 1 with metastases to both liver and lung) (Tables 2 and 4). One of these tumors was grade 1, 2 were grade 2, and 2 were grade 3. Two of these patients, both with unilateral pure carcinomas, had PP that was described intraoperatively as mucinous ascites with multiple small mucinous tumor implants in the abdomen and pelvis; in one case there was a large omental mucinous tumor. Biopsy of the extra-ovarian tumor in these two cases revealed metastatic carcinoma with malignant epithelial cells lying in large pools of mucin that had dissected into various tissues. One of these patients had a normal barium enema preoperatively, but the appendix and no other portion of the gastrointestinal tract was removed or commented on in the operative report in either of the cases.

Back to Top | Article Outline
Summary of Cases With Rupture

Among the stage I cases with follow-up data, there were 13 with either preoperative or intraoperative tumor rupture[colon] 4 in the borderline with atypia category, 6 in the borderline-with-intraepithelial carcinoma category, and 3 in the invasive carcinoma category. One (8[percnt]) of the 13 patients had a recurrence and died, the 18-year-old woman with a borderline tumor with intraepithelial carcinoma (Table 4). Two of 125 (1.6[percnt]) patients whose stage I tumors had not ruptured had persistence or recurrence of the tumor.

Back to Top | Article Outline
Summary of Cases With Pseudomyxoma Ovarii

There were 28 cases with PO and follow-up data, 11 (39[percnt]) of which were accompanied by PP; 18 of the 28 cases were in the borderline with atypia category, 7 in the borderline with intraepithelial carcinoma category, and 3 in the invasive carcinoma category. The PO contained epithelial cells in 8, 1, and 3 cases in these categories, respectively.

Of the 16 tumors characterized by PO without epithelial cells, one recurred, the borderline tumor with intraepithelial carcinoma with intraoperative rupture discussed above (Tables 2 and 4). Of the 12 tumors in which the PO contained epithelial cells, 6 recurred or persisted (4 borderline with atypia, 2 invasive). All 6 were stage III with intraoperative evidence of PP, 4 of the dissecting with fibrosis type with atypical cells in the mucin, 1 of unknown type, and 2 with metastatic mucinous carcinoma (Tables 3 and 4). Of the 6 cases with PO containing epithelial cells in which there was no recurrence, 4 were in the borderline (with atypia) category (1 stage Ia, 1 stage Ic with PP of the superficial organizing type, 2 stage III, one with PP of the dissecting-with-fibrosis type, and the other with PP of unknown type (Table 3), 1 in the borderline with intraepithelial carcinoma category (stage Ia), and one stage IIa infiltrative carcinoma.

Back to Top | Article Outline


This study confirms the observations of others that the prognosis of ovarian mucinous borderline tumors of the intestinal type and mucinous carcinomas is highly dependent on their stage. For the three major categories of tumor (borderline, borderline with intraepithelial carcinoma, and invasive carcinoma) combined, the 5-year disease-free survival rate in stage I cases was 98[percnt] ([plusmn]1.0[percnt]) in contrast to 35[percnt] ([plusmn]14[percnt]) for stages II through IV (p [le]0.001). The cumulative survival figures from the literature are 95[percnt] for stage I cases in all three categories combined (1103 of 1155) 1,5[ndash]8,12,13,15,16,18,20,23,24,29,31,38,39,44,47,48,56,59,61[ndash]65 and 32[percnt] for higher-stage cases (42 of 132). 6,15,18,20,23,24,27,31,39,56,59,64 Notably, all three tumors considered stage I in this series that recurred were incompletely staged.

This study and those cited above have also shown that the stage and the type of extra-ovarian spread are related to the histologic features of the ovarian tumor. In the current study, all seven FIGO stage III tumors in the borderline (with epithelial atypia) category presented with intraoperative evidence of [ldquo]pseudomyxoma peritonei.[rdquo] All 90 tumors in the borderline with intraepithelial carcinoma category were stage I, but two of the 69 with follow-up data recurred, both with invasive abdominal metastases without PP. Seven of 32 (22[percnt]) tumors in the invasive (including microinvasive) carcinoma category were at an advanced stage at presentation, all seven had infiltrative stromal invasion greater than microinvasive, and all of the peritoneal lesions in these cases were also infiltrative (two with clinical PP and abundant extracellular mucin in the implants).

Except for four tumors that were uniformly carcinomatous, malignant mucinous ovarian tumors formed a morphologic continuum with foci of benign-appearing tumor, borderline tumor, or both present in specimens with intraepithelial and invasive carcinomas. Molecular biologic studies have revealed the lowest frequency of K-ras mutations in cystadenomas, an intermediate frequency in borderline tumors, and the highest frequency in mucinous carcinomas. 19,26,28 These two lines of evidence suggest the possibility of progression from benign to malignant neoplasia in the evolution of at least some mucinous carcinomas, as postulated by others, 36,39 and are in contrast to the observation that fewer serous carcinomas contain benign and borderline components. 2,36,50 A third piece of evidence, based on cumulative figures from the literature, is that the average age of 795 women with benign mucinous tumors was 42.6 years; of 417 women with borderline mucinous tumors, 46.8 years; and of 690 women with mucinous carcinomas, 52.8 years. 49 A similar difference in age incidence, however, did not emanate from analysis of our relatively small series of cases in that patients with borderline (with atypia) tumors, borderline tumors with intraepithelial carcinoma, and invasive carcinomas all had similar average ages.

Back to Top | Article Outline
Borderline Tumors (With Atypia)

The morphologic continuum in mucinous tumors causes difficulty in their histopathologic subclassification. 37 Moreover, although the WHO recommended that proliferating mucinous tumors without evidence of stromal invasion be considered borderline, 53,54 Hart and Norris 16 and later Hart 17 removed tumors that appeared noninvasive but had epithelial stratification greater than three layers (devoid of stromal support and generally with moderate to marked nuclear atypia) from the borderline category and designated them carcinomas. They did so because 3 of 12 (25[percnt]) patients with stage I tumors with those features died of disease in contrast to only 3 of 87 (3[percnt]) patients with stage I tumors that lacked those features. The latter tumors were retained in the borderline category. Subsequent investigators who adopted the Hart-Norris criteria for mucinous borderline tumors have reported fatal recurrences in only 2 of 400 (0.5[percnt]) stage I cases. 1,5[ndash]8,12,13,15,38,39,56,57,61,64 In contrast, in studies of mucinous borderline tumors in which the diagnostic criteria were not specified and which undoubtedly included some tumors in the Hart-Norris (noninvasive) carcinoma group (our borderline with intraepithelial carcinoma category), recurrences were reported in 15 of 439 (3.4[percnt]) stage I cases. 20,24,29,31,44,58,61 Recurrences in series composed entirely of noninvasive mucinous carcinomas of the Hart-Norris variety have been reported in 10 of 139 (7.2[percnt]) stage I cases. 7,12,15,16,18,23,39,56,57,63 Almost all of the recurrent tumors with adequate follow-up data in the latter series of cases were fatal. Hart and Norris acknowledged that all three stage I tumors in their series that were borderline according to their criteria and had adverse outcomes had been incompletely sampled, possibly leading to underdiagnosis of foci of intraepithelial or even invasive carcinoma. In the only other report of stage I borderline tumors with atypia that included fatal cases (two), 7 sampling data are not given. Thus, stage I tumors in the borderline (with atypia) category, except for those in which sampling may have been inadequate, have had a benign behavior, as confirmed in this study. Because of this finding, some investigators have advocated eliminating the [ldquo]borderline[rdquo] designation for these tumors and replacing it with [ldquo]proliferating[rdquo] or [ldquo]atypical proliferating.[rdquo]5,39,41,44,55 We favor the retention of [ldquo]borderline,[rdquo] however, because of the difficulty in separating some of them from borderline tumors with intraepithelial carcinoma, which have been shown to be capable of malignant behavior; the possibility that occult intraepithelial or invasive carcinoma exists in them despite what is even considered adequate sampling; and the possibility that mucinous borderline tumors are at least a rare source of PP of the dissecting-with-fibrosis type (see below).

Back to Top | Article Outline
Relation of Borderline Tumors (With Atypia) to `Pseudomyxoma Peritonei'

Nine of the 74 borderline tumors (with atypia) in our series (12[percnt]) and a similar proportion in the literature have been associated with [ldquo]pseudomyxoma peritonei,[rdquo] a disorder enshrouded in controversy. A major problem with PP is the lack of a uniformly accepted definition. Although the term has been applied most often to a chronic accumulation of intraabdominal mucus that is generally associated with a slowly growing mucinous neoplasm, the designation has also been used for cases ranging from metastatic carcinomas on the peritoneum that secrete large amounts of mucin and have a relatively rapid clinical course to cases in which an acute or recent rupture of an ovarian mucinous tumor of any type has resulted in the presence of free-floating or organizing mucus in the abdominal cavity. We agree with Ronnett et al. 43 that PP should be eliminated as a pathologic term. Instead of their subdivision of intraabdominal mucus into two types designated by them [ldquo]diffuse peritoneal adenomucinosis[rdquo] and [ldquo]peritoneal mucinous carcinomatosis,[rdquo] however, we prefer to subdivide this disorder when it is associated with ovarian cystic tumors into three categories[colon] organizing intraabdominal mucin that is superficially attached to peritoneal surfaces (organizing mucinous ascites), noncarcinomatous dissecting mucin with reactive fibrosis, and metastatic mucinous carcinoma (with or without dissection and fibrosis). In each of the first two categories, we state whether epithelial cells are found within the pools of mucin and, if so, whether they have benign or borderline (atypical) nuclear features, as described in the Materials and Methods section. We know of no evidence that the first category of intraabdominal mucin progresses to the second, the type associated with chronic, usually progressive disease.

Another problem with PP relates to the primary or secondary nature of the ovarian tumors that have been classified in this series and in most reported series as FIGO stage II and stage III ovarian mucinous borderline tumors (of intestinal type). In reported studies of cases categorized as such, the extra-ovarian spread has been in the form of PP in 48 of 49 reported cases 6,8,15,24,27,29,45,56 and in the present series, 7 of 7 cases (Table 3). Most authors over the past several decades have considered tumors of this type primary in the ovary and have reported them as stage II and III ovarian tumors according to the staging systems of FIGO, the International Union Against Cancer, and the American Joint Commission on Cancer. Several groups of authors, particularly those writing in the last decade, however, have reinterpreted such tumors as metastatic from the appendix (in some, most, or all cases). 18,35,41,42,65

A third problem with PP is caused by a lack of data in the literature to establish the frequency of concomitant appendiceal lesions of similar histologic type in patients with ovarian mucinous cystic tumors and PP of the dissecting-with-fibrosis type. A valid figure for this association cannot be given because in many cases the appendix has not been inspected at operation or removed surgically; if removed, it has not been examined microscopically; or if examined microscopically, it has not been processed by the step-sectioning that may be necessary to detect a microscopic focus of tumor with evidence of rupture. In a review of several recently reported series in which microscopic examination of the appendix of unknown thoroughness has been performed in cases of ovarian mucinous cystic tumors accompanied by PP, the appendix was reported to contain a mucinous tumor in approximately two thirds of the cases, 10,20,21,27,35,41 but a small tumor may have been overlooked in the other cases as a result of inadequate sampling.

A final problem, in cases in which an appendiceal tumor has been identified, is a lack of consensus among pathologists whether the accompanying cystic ovarian tumor is metastatic or independently primary. Several papers based on clinicopathologic, immunohistochemical, and molecular genetic research have appeared with respect to this controversy, with most, but not all of them, supporting the metastatic nature of the ovarian tumors in this situation. 9,11,14,35,41,42,52,58,65 We think that of the seven tumors that we classified according to FIGO criteria as stage III ovarian borderline tumors, five, three of which were documented pathologically to be of the dissecting-with-fibrosis type, were probably metastatic from the appendix because of bilaterality in all five cases, grossly observable ovarian surface involvement in three of them, and a microscopically similar tumor in the single resected appendix; the appendix was not sampled or mentioned in the operative report in the other four cases. In a sixth case in which the PP was probably of the dissecting-with-fibrosis type, on the other hand, the ovarian tumor was probably primary because it was unilateral and associated with a dermoid cyst. In the final stage III case, in which the PP was of the dissecting-with-fibrosis type, the ovarian tumor was unilateral and left-sided, and despite failure to remove the appendix (which was not commented on in the operative report), was associated with disease-free survival 8 years postoperatively, suggesting but not establishing an ovarian origin.

Although there is compelling evidence that in most cases PP of the dissecting-with-fibrosis type associated with ovarian mucinous cystic [ldquo]borderline tumors of intestinal type[rdquo] does not originate from a primary ovarian neoplasm, we have retained the standard approach of staging such tumors as primary ovarian tumors because a consensus has not developed on criteria for distinguishing primary ovarian tumors with PP from those of appendiceal origin. The current staging approach is already well-established among gynecologic oncologists, and abolishing it or altering it significantly at the present time without a consensus would lead to irreproducibility of staging. Also, there is currently no internationally or nationally sanctioned system for staging appendiceal tumors. Based partly on the accumulating data that the ovarian tumors are metastatic in cases in which an appendiceal mucinous lesion is also present, we propose criteria for determining the ovarian or appendiceal origin of PP associated with ovarian mucinous cystic tumors in Table 5, and urge staging committees to alter their rules for tumors associated with PP.

Back to Top | Article Outline
Borderline Tumors With Intraepithelial Carcinoma

This study confirms the usefulness of separating tumors within the WHO borderline category into two subtypes, as was advocated by Hart and Norris, who introduced different terminology than we propose. 16 Including two cases in the current study, 5.8[percnt] (12 of 208) of reported stage I tumors in the borderline with intraepithelial carcinoma category have recurred. 7,12,15,16,18,23,39,56,57,63 This finding contrasts with the negligible recurrence rate of stage I tumors in the borderline (with atypia) category. Also, in contrast to the invasive carcinomas in our series, 22[percnt] of which were of high stage, no borderline-with-intraepithelial carcinomas were higher than stage I (although 1 of the 2 tumors that recurred was stage Ic because of positive peritoneal washings and both recurrent tumors were incompletely staged). In prior reports, 10 of 139 (7.2[percnt]) borderline tumors with intraepithelial carcinoma have been stage II or III; and 6 of the 10 tumors with follow-up data have been fatal. 7,12,15,16,18,23,39,56,57,63 It is likely, however, that in most, if not all, of the higher-stage tumors and stage I tumors that recurred, foci of invasion in the primary tumor had not been sampled.

Unfortunately, the use of several different terms for tumors with malignant architectural and nuclear features that lack clearly identifiable stromal invasion has led to confusion in classifying ovarian mucinous tumors. Other investigators have designated these tumors [ldquo]noninvasive mucinous carcinomas,[rdquo]7,15,16,63 [ldquo]well-differentiated or grade I mucinous carcinomas,[rdquo]5,12 [ldquo]intraglandular carcinomas,[rdquo]18 [ldquo]grade 2, 3 or 4 borderline tumors,[rdquo]56,57 or [ldquo]atypical proliferative mucinous tumors with intraepithelial carcinoma.[rdquo]39 We prefer the term [ldquo]borderline with intraepithelial carcinoma[rdquo] because it retains the long-established borderline terminology yet identifies both an atypical and a frankly malignant-appearing component, with the latter indicating a small risk of adverse clinical behavior. We have included cases in this category that others may have retained in the borderline (with atypia) category, 15,18 because we did not require a minimum quantity of malignant-appearing foci or [ldquo]marked[rdquo] nuclear atypia for inclusion. This approach seems valid because none of our stage I tumors in the borderline category recurred, and the two tumors that recurred in the borderline-with-intraepithelial-carcinoma category had only minor (5[percnt] or less) areas that qualified them for this designation.

Some authors' data suggest that recurrences in stage I tumors in this category can be predicted by grading the malignant intraepithelial component 44,57 or by ploidy analysis. 15 In the two fatal cases from this study, there were foci, albeit small, of grade 3 nuclei, but 12 other stage I tumors with follow-up data in this category that contained similar, usually larger, foci did not recur. In another study, all 12 patients with [ldquo]grade 3[rdquo] borderline tumors survived. 56 Therefore, although the presence of grade 3 nuclear atypia may warrant a more vigorous search for invasive foci, it is doubtful that grading of well-sampled mucinous borderline tumors with intraepithelial carcinoma is prognostically useful. Other investigators have drawn similar conclusions. 63

Finally, the absence of bilaterality or associated PP in this series of 90 cases of mucinous borderline tumors with intraepithelial carcinoma reinforces the conclusions that almost all bilateral ovarian mucinous tumors and tumors associated with PP containing benign-appearing or borderline epithelial cells do not belong in the primary ovarian tumor category. If they did, one would expect to find foci of intraepithelial carcinoma in over half of such cases according to our percentage distribution of cases in the borderline (with atypia) and borderline with intraepithelial carcinoma categories.

Back to Top | Article Outline
Invasive Carcinomas Including Microinvasive Forms

Only one of the 20 (5.0[percnt]) stage I mucinous carcinomas with follow-up data in our study recurred. In other series reported since 1990, 11 of 96 (11.5[percnt]) stage I invasive mucinous carcinomas have recurred. 15,18,23,39,47,48,63 None of the five carcinomas with only microinvasion recurred. Although the definition of what constitutes microinvasion has varied, this favorable outcome is in accordance with an expanding number of similar observations 15,18,22,30,39 and is similar to the current reported experience with stage I serous borderline tumors with microinvasion. 3,60 However, Watkin et al. 63 reported that 6 of the 12 mucinous carcinomas with stromal invasion that they studied, 10 of which were fatal, had only [ldquo]rare microscopic foci[rdquo] of invasion (not further defined). Most, possibly all, of these cases were stage III, however.

None of the 12 mucinous carcinomas with purely expansile invasion had spread beyond the ovary and all 10 patients with follow-up data survived without recurrence. Notably, all of the tumors in this category were diagnosed as carcinoma by the original pathologists, indicating that they are generally recognized as cancers. Hoerl and Hart 18 allude to the [ldquo]difficulty in identifying patterns of unequivocal stromal invasion in some otherwise obviously malignant ovarian carcinomas.[rdquo] Our recognition of an expansile subcategory and the [ldquo]confluent invasive[rdquo] patterns described by Riopel et al. 39 reflect, in part, this difficulty. Some of the tumors that we have included in this category, particularly those with closely approximated, but not back to back, malignant glands (Fig. 7), may have been considered [ldquo]noninvasive[rdquo] or [ldquo]intraglandular[rdquo] carcinomas by others 15,18,63 who have also demonstrated a favorable outcome for patients with such tumors.

The findings in the present series indicate that infiltrative stromal invasion is biologically more aggressive than expansile invasion. The only stage I carcinoma that recurred had infiltrative invasion. Also, seven of the 20 carcinomas with infiltrative invasion (including microinvasion) had spread beyond the ovary by the time of presentation; and 5 of the 6 (83[percnt]) tumors with follow-up data were fatal (Tables 2 and 4). It is possible, however, that in both of the fatal cases with PP, the tumors were metastatic from the appendix or elsewhere in the intestine. In other studies of mucinous ovarian carcinomas with extra-ovarian spread, almost all of them have also exhibited infiltrative invasion in the ovarian tumor, and 94[percnt] (30 of 32 cases) have been fatal. 15,18,23,63

Back to Top | Article Outline
Tumor Rupture

Including the cases in the current series and those previously reported, 41 stage I tumors in the borderline (with atypia) category had ruptured preoperatively or intraoperatively. 15,16,24 None of them have recurred or led to the development of PP, however, lending additional support to the conclusion that so-called borderline ovarian mucinous tumors with PP are most often metastatic.

In contrast, however, three of 20 (15[percnt]) stage I tumors in the borderline with intraepithelial carcinoma category 15,16 and 2 of 7 (29[percnt]) in the carcinoma category 16,47 with rupture in this and previously reported series have recurred. These rates are higher than the reported recurrence rates for all stage I tumors in these categories (5.8[percnt] and 11[percnt]) and suggest an adverse effect resulting from rupture in these categories, although more cases need to be studied to confirm this impression.

Back to Top | Article Outline
Pseudomyxoma Ovarii

Michael et al. 27 reported 11 cases of ovarian mucinous tumors associated with PP with only atypia of their neoplastic cells but with extensive PO. The ovarian pools of mucin contained epithelial cells in most of the cases, and the PP contained epithelial cells in all of them. These authors concluded that the presence of PO with epithelial cells in tumors that otherwise appear to be of low malignant potential (borderline) indicates stromal invasion with a potential for peritoneal spread and justifies the designation [ldquo]carcinoma.[rdquo] The authors of two subsequent studies of ovarian mucinous tumors with PP have drawn a similar conclusion. 15,21 In these three studies, however, as in the five cases in the current study with similar findings, most of the ovarian tumors may have been metastatic from an appendiceal mucinous tumor. In the present study, there were two stage I tumors in the borderline with atypia category and one in the borderline with intraepithelial carcinoma category with PO containing cells; none of these tumors recurred in follow-up periods of 2 to 9 years. Because PO with or without cells may be initiated by mucin extravasation into the stroma rather than true tumor cell invasion 65 and stage I tumors with PO containing atypical cells have not been reported to recur, the designation of such tumors as invasive carcinoma does not appear justified on the basis of current information.

Back to Top | Article Outline


Malignant mucinous ovarian tumors present unique problems among surface epithelial-stromal cancers of the ovary. These include tumor heterogeneity in many cases requiring judicious, and at times extensive sampling for microscopic evaluation; variations in the terminology used for subsets of these tumors, making comparisons from one series to another difficult; problems in the differential diagnosis of primary and metastatic mucinous ovarian tumors, and differing opinions about the nature and terminology of [ldquo]pseudomyxoma peritoneii.[rdquo]

This retrospective study of 196 cases of borderline and carcinomatous mucinous ovarian tumors was undertaken in an attempt to resolve the above problems, but, like most other such series, was beset by inadequate surgical staging and pathologic sampling in the small number of cases in which the subsequent fate of the patient was unexpected, and in most of the cases in which the puzzling complication PP was present. It was necessary, therefore, to present the problem cases in this series in some detail to enable the readers to understand their complexity, provide surgeons and pathologists information about potential pitfalls in their evaluation of these tumors, and alert them to the necessity of more detailed investigations in the future.

On the basis of the present study and other studies of ovarian mucinous borderline tumors and carcinomas, however, the following conclusions can be drawn.

If one excludes cystadenomas and endocervical-like mucinous tumors, the latter accounting for 5[percnt] to 14[percnt] of mucinous borderline tumors, mucinous ovarian tumors can be separated on the basis of pathologic and prognostic differences into three major categories[colon] borderline (with atypia), borderline with intraepithelial carcinoma, and invasive carcinoma. Stage I tumors in all three categories are almost always unilateral. The risk of recurrence in the first category is extremely small, with a low but progressively increasing risk in the second and third categories. Unilateral oophorectomy in stage Ia cases, especially in the first two categories, is associated with minimal risk. Preoperative or intraoperative tumor rupture in stage I cases in the borderline (with atypia category) does not appear to predispose to recurrence. However, there may be a slightly increased risk in the other two categories.

Extra-ovarian disease associated with tumors that have been customarily classified in the borderline (with atypia) category is almost always in the form of [ldquo]pseudomyxoma peritonei,[rdquo] an imprecise clinical or surgical term that should be replaced by specific histologic terminology that is based on microscopic analysis of generously sampled abdominal deposits. In view of the emerging consensus that most, if not all, ovarian mucinous cystic tumors associated with PP are metastatic from the appendix, or less often another gastrointestinal source, we recommend that staging committees such as those of FIGO, UICC, and AJCC, reconsider their rules for staging of such tumors to reflect the evolving knowledge about them. We have proposed a set of criteria for determining the primary or metastatic nature of these ovarian neoplasms for consideration by staging committees in Table 5.

Some ovarian mucinous tumors associated with PP have a component of dermoid cyst. Although more cases of this type need to be studied, the PP in these cases appears more likely to be of the superficial organizing type, which is associated with a favorable prognosis.

The great majority of the ovarian tumors associated with PP (including all of the cases in our series) also have [ldquo]pseudomyxoma ovarii,[rdquo] with atypical epithelium in the ovarian mucin in most of the cases. Although the number of cases is small, stage I tumors with PO but without PP have not been shown to recur, suggesting that these tumors should not be considered carcinomas on the basis of the PO alone.

Tumors in the borderline with intraepithelial carcinoma category are almost always stage Ia but may present at a higher stage. The risk of recurrence in stage I cases, including those in the current series, has been 5.8[percnt]. In such cases, as in cases presenting at a higher stage, it is highly probable that occult areas of invasion in the primary tumor were missed because of inadequate sampling. The presence of high-grade nuclear atypia may warrant additional sampling. However, in our series of cases, recurrence of stage I tumors with intraepithelial carcinoma could not be predicted by either grading or quantifying the intraepithelial carcinomatous component of these tumors. Although experience is still limited, the additional finding of microinvasion does not appear to affect outcome.

Including cases in the current series, stage I mucinous carcinomas reported since 1990 have been almost always unilateral and have had a recurrence rate of 11.5[percnt]. It is possible that earlier series with significantly higher recurrence rates have included cases that had either been understaged or were metastatic to the ovary from an occult primary mucinous carcinoma. Mucinous carcinomas with infiltrative invasion are more likely to present at a higher stage than those with expansile invasion, indicating that the former pattern reflects an enhanced ability to spread. Grading of stage I ovarian mucinous carcinomas was not predictive of their potential to recur in our series and other series of cases. 18,63 However, higher-grade tumors are more likely to present at a higher stage. In almost all cases, peritoneal implants of primary ovarian mucinous carcinomas are invasive without PP; the presence of the latter strongly suggesting a mucinous carcinoma metastatic to the ovary. Extra-ovarian spread of ovarian mucinous carcinomas is almost always fatal.

Many of the conclusions about malignant mucinous tumors of the ovary have been based on relatively few cases in each subcategory, and have included cases in which staging or microscopic sampling or both have been suboptimal. Only with more detailed future investigations will the many uncertainties associated with these tumors be satisfactorily reduced or eliminated.

Back to Top | Article Outline


The authors thank Kathleen P. Lee, PhD, for invaluable assistance with data acquisition and computer entry, Elizabeth E. Allred, MS, for statistical assistance, and Kalimah Nadeem-Holts for typing the manuscript.

Back to Top | Article Outline


1. Barnhill D, Heller P, Brzozowski P, et al. Epithelial ovarian carcinoma of low malignant potential. Obstet Gynecol 1985; 65[colon]53[ndash]9.
2. Bell DA, Scully RE. Early de novo ovarian carcinoma[colon] a study of fourteen cases. Cancer 1994; 73[colon]1859[ndash]64.
3. Bell DA, Scully RE. Ovarian serous borderline tumors with stromal microinvasion[colon] a report of 21 cases. Hum Pathol 1990; 21[colon]397[ndash]403.
4. Berezowski K, Stastny JF, Kornstein MJ. Cytokeratins 7 and 20 carcinoembryonic antigen in ovarian and colonic carcinoma. Mod Pathol 1996; 9[colon]426[ndash]9.
5. Bostwick DG, Tazelaar HD, Ballon SC, et al. Ovarian epithelial tumors of borderline malignancy[colon] a clinical and pathologic study of 109 cases. Cancer 1986; 58[colon]2052[ndash]65.
6. Casey AC, Bell DA, Lage JM, et al. Epithelial ovarian tumors of borderline malignancy[colon] long-term follow-up. Gynecol Oncol 1993; 50[colon]316[ndash]22.
7. Chaitin BA, Gershenson DM, Evans HL. Mucinous tumors of the ovary[colon] a clinicopathologic study of 70 cases. Cancer 1985; 55[colon]1958[ndash]62.
8. Chambers JT, Merino MJ, Kohorn EI, et al. Borderline ovarian tumors. Am J Obstet Gynecol 1988; 159[colon]1088[ndash]94.
9. Chuaqui RF, Zhuang Z, Emmert-Buck MR, et al. Genetic analysis of synchronous mucinous tumors of the ovary and appendix. Hum Pathol 1996; 27[colon]165[ndash]71.
10. Costa MJ. Pseudomyxoma peritonei[colon] histologic predictors of patient survival. Arch Pathol Lab Med 1994; 118[colon]1215[ndash]9.
11. Cuatrecasas M, Matias-Guiu X, Prat J. Synchronous mucinous tumors of the appendix and the ovary associated with pseudomyxoma peritonei[colon] a clinicopathologic study of six cases with comparative analysis of c-Ki-ras mutations. Am J Surg Pathol 1996; 20[colon]739[ndash]46.
12. De Nictolis M, Montironi R, Tommasoni S, et al. Benign, borderline and well-differentiated malignant intestinal mucinous tumors of the ovary[colon] a clinicopathologic, histochemical, immunohistochemical, and nuclear quantitative study of 57 cases. Int J Gynecol Pathol 1994; 13[colon]10[ndash]21.
13. Fort MG, Pierce VK, Saigo PE, et al. Evidence for the efficacy of adjuvant therapy in epithelial ovarian tumors of low malignant potential. Gynecol Oncol 1989; 32[colon]269[ndash]72.
14. Guerrieri C, Fr[aring]nlund B, Boeryd B. Methods in pathology[colon] expression of cytokeratin 7 in simultaneous mucinous tumors of the ovary and appendix. Mod Pathol 1995; 8[colon]573[ndash]6.
15. Guerrieri C, H[ouml]gberg T, Wingren S, et al. Mucinous borderline and malignant tumors of the ovary[colon] a clinicopathologic and DNA ploidy study of 92 cases. Cancer 1994; 74[colon]2329[ndash]40.
16. Hart WR, Norris HJ. Borderline and malignant mucinous tumors of the ovary[colon] histologic criteria and clinical behavior. Cancer 1973; 31[colon]1031[ndash]45.
17. Hart WR. Ovarian epithelial tumors of borderline malignancy (carcinomas of low malignant potential). Hum Pathol 1977; 8[colon]541[ndash]9.
18. Hoerl HD, Hart WR. Primary ovarian mucinous cystadenocarcinomas. A clinicopathologic study of 49 cases with long-term follow-up. Am J Surg Pathol 1998; 22[colon]1449[ndash]62.
19. Ichikawa Y, Nishida M, Suzuki H, et al. Mutation of K-ras protooncogene is associated with histological subtypes in human mucinous ovarian tumors. Cancer Res 1994; 54[colon]33[ndash]5.
20. Kaern J, Trop[eacute] CG, Abeler VM. A retrospective study of 370 borderline tumors of the ovary treated at the Norwegian Radium Hospital from 1970 to 1982. Cancer 1993; 71[colon]1810[ndash]20.
21. Kahn MS, Demopoulos RI. Mucinous ovarian tumors with pseudomyxoma peritonei[colon] a clinicopathological study. Int J Gynecol Pathol 1992; 11[colon]15[ndash]23.
22. Khunamornpong S, Russell P, Dalrymple JC. Proliferating (LMP) mucinous tumors of the ovaries with microinvasion[colon] morphologic assessment of 13 cases. Int J Gynecol Pathol 1999; 18[colon]238[ndash]46.
23. Kikkawa F, Kawai M, Tamakosh K, et al. Mucinous carcinoma of the ovary. Clinicopathologic analysis. Oncology 1996; 53[colon]303[ndash]7.
24. Kliman L, Rome RM, Fortune DW. Low malignant potential tumors of the ovary[colon] a study of 76 cases. Obstet Gynecol 1986; 68[colon]338[ndash]44.
25. Loy TS, Calaluce RD, Keeney GL. Cytokeratin immunostaining in differentiating primary ovarian carcinoma from metastatic colonic adenocarcinoma. Mod Pathol 1996; 9[colon]1040[ndash]4.
26. Mandai M, Konishi I, Kuroda H, et al. Heterogeneous distribution of K-ras-mutated epithelia in mucinous ovarian tumors with special reference to histopathology. Hum Pathol 1998; 28[colon]34[ndash]40.
27. Michael H, Sutton G, Roth LM. Ovarian carcinoma with extracellular mucin production[colon] reassessment of `pseudomyxoma ovarii et peritonei.' Int J Gynecol Pathol 1987; 6[colon]298[ndash]312.
28. Mok SC-H, Bell DA, Knapp RC, et al. Mutation of K-ras protooncogene in human ovarian epithelial tumors of borderline malignancy. Cancer Res 1993; 53[colon]1489[ndash]92.
29. Nakashima N, Nagasaka T, Oiwa N, et al. Ovarian epithelial tumors of borderline malignancy in Japan. Gynecol Oncol 1990; 38[colon]90[ndash]8.
30. Nayar R, Siriaunkgul S, Robbins KM, et al. Microinvasion in low malignant potential tumors in the ovary. Hum Pathol 1996; 27[colon]521[ndash]7.
31. Nikrui N. Survey of clinical behavior of patients with borderline epithelial tumors of the ovary. Gynecol Oncol 1981; 12[colon]107[ndash]19.
32. Pecorelli S, Odicino F, Maisonneuve P, et al. Carcinoma of the ovary. In[colon] FIGO[colon] Annual reports on the results of treatment in gynecological cancer. J Epidemiol Biostat 1998;3[colon]75[ndash]102.
33. Prat J, Scully RE. Sarcomas in ovarian mucinous tumors[colon] a report of two cases. Cancer 1979; 44[colon]1327[ndash]31.
34. Prat J, Young RH, Scully RE. Ovarian mucinous tumors with foci of anaplastic carcinoma. Cancer 1982; 50[colon]300[ndash]4.
35. Prayson RA, Hart WR, Petras RE. Pseudomyxoma peritonei[colon] a clinicopathologic study of 19 cases with emphasis on site of origin and nature of associated ovarian tumors. Am J Surg Pathol 1994; 18[colon]591[ndash]603.
36. Puls LE, Powell DE, DePriest PD, et al. Transition from benign to malignant epithelium in mucinous and serous ovarian cystadenocarcinoma. Gynecol Oncol 1992; 47[colon]53[ndash]7.
37. Raab SS, Robinson RA, Jensen CS, et al. Mucinous tumors of the ovary[colon] interobserver diagnostic variability and utility of sectioning protocols. Arch Pathol Lab Med 1997; 121[colon]1192[ndash]8.
38. Rice LW, Berkowitz RS, et al. Epithelial ovarian tumors of borderline malignancy. Gynecol Oncol 1990; 39[colon]195[ndash]8.
39. 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[colon]617[ndash]35.
40. Ronnett BM, Kurman RJ, Shmookler BM, et al. The morphologic spectrum of ovarian metastases of appendiceal adenocarcinomas[colon] 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[colon]1144[ndash]55.
41. Ronnett BM, Kurman RJ, Zahn CM, et al. Pseudomyxoma peritonei in women[colon] a clinicopathologic analysis of 30 cases with emphasis on site of origin, prognosis, and relationship to ovarian mucinous tumors of low malignant potential. Hum Pathol 1995; 26[colon]509[ndash]24.
42. Ronnett BM, Shmookler BM, Diener-West M, et al. Immunohistochemical evidence supporting the appendiceal origin of pseudomyxoma peritonei in women. Int J Gynecol Pathol 1997; 16[colon]1[ndash]9.
43. Ronnett BM, Zahn CM, Kurman RJ, et al. Disseminated peritoneal adenomucinosis and peritoneal mucinous carcinomatosis[colon] 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[colon]1390[ndash]408.
44. Russell P, Merkur H. Proliferative ovarian `epithelial' tumors[colon] a clinicopathological analysis of 144 cases. Aust NZJ Obstet Gynecol 1979; 19[colon]45[ndash]51.
45. Russell P. The pathological assessment of ovarian neoplasms. In[colon] The proliferating `epithelial' tumors. Pathology 1979;11[colon]251[ndash]82.
46. Rutgers JL, Scully RE. Ovarian mullerian mucinous papillary cystadenomas of borderline malignancy[colon] a clinicopathologic analysis. Cancer 1988; 61[colon]340[ndash]8.
47. Schueler JA, Cornelisse CJ, Hermans J, et al. Prognostic factors in well-differentiated early-stage epithelial ovarian cancer. Cancer 1993; 71[colon]787[ndash]95.
48. Schueler JA, Trimbos JB, Burg M, et al. DNA index reflects the biological behavior of ovarian carcinoma stage I[ndash]IIa. Gynecol Oncol 1996; 62[colon]59[ndash]66.
49. Scully RE. Ovary. In[colon] Henson ED, Albores-Saavedra J, eds. Pathology of Incipient Neoplasia, 2nd ed. Philadelphia, PA[colon] WB Saunders, 1993[colon]289[ndash]90.
50. Scully RE, Bell DA, Abu-Jawdeh GM. Update on early ovarian cancer and cancer developing in benign ovarian tumors. In[colon] Mason P, Sharp F, Blackett A, et al., eds. Ovarian Cancer, Biological and Therapeutic Challenges. London[colon] Chapman and Hall, 1995[colon]139[ndash]44.
51. Scully RE, Young RH, Clement PB. Tumors of the ovary, maldeveloped gonads, fallopian tube and broad ligament. In[colon]Atlas of Tumor Pathology, 3rd series, fascicle 23. Washington, DC[colon] Armed Forces Institute of Pathology, 1998.
52. Seidman JD, Elsayed AM, Sobin LH, et al. Association of mucinous tumors of the ovary and appendix[colon] a clinicopathologic study of 25 cases. Am J Surg Pathol 1993; 17[colon]22[ndash]34.
53. Serov SF, Scully RE, Sobin LH. International histological classification of tumors (no. 9). Histological Typing of Ovarian Tumours. Geneva[colon] World Health Organization, 1973.
54. Scully RE, World Health Organization. International histological classification of tumors. Histological Typing of Ovarian Tumours. Heidelberg[colon] Springer, 1999.
55. Silva EG, Kurman RJ, Russell P, et al. Symposium[colon] ovarian tumors of borderline malignancy. Int J Gynecol Pathol 1996; 15[colon]281[ndash]302.
56. Siriaunkgul S, Robbins KM, McGowan L, et al. Ovarian mucinous tumors of low malignant potential[colon] a clinicopathologic study of 54 tumors of intestinal and mullerian type. Int J Gynecol Pathol 1995; 14[colon]198[ndash]208.
57. Sumithran E, Susil BJ, Looi LM. The prognostic significance of grading in borderline mucinous tumors of the ovary. Hum Pathol 1988; 19[colon]15[ndash]8.
58. Sumithran E, Susil BJ. Concomitant mucinous tumors of appendix and ovary. Cancer 1992; 70[colon]2980[ndash]3.
59. Tasker M, Langley F. The outlook for women with borderline epithelial tumors of the ovary. Br J Obstet Gyncol 1985; 92[colon]969[ndash]73.
60. Tavassoli FA. Serous tumor of low malignant potential with early stromal invasion (serous LMP with microinvasion). Mod Pathol 1988; 1[colon]407[ndash]14.
61. Tazelaar HD, Bostwick DG, Ballon SC, et al. Conservative treatment of borderline ovarian tumors. Obstet Gynecol 1985; 66[colon]417[ndash]22.
62. Trope C, Kaern J, Vergot IB, et al. Are borderline tumors of the ovary overtreated both surgically and systemically? A review of four prospective randomized trials including 253 patient with borderline tumors. Gynecol Oncol 1993; 51[colon]236[ndash]43.
63. Watkin W, Silva EG, Gershenson DM. Mucinous carcinoma of the ovary[colon] pathologic prognostic factors. Cancer 1992; 69[colon]208[ndash]12.
64. Yoonessi M, Crickard K, Celik C, et al. Borderline epithelial tumors of the ovary[colon] ovarian intraepithelial neoplasia. Obstet Gynecol Surv 1988; 43[colon]435[ndash]44.
65. Young RH, Gilks CB, Scully RE. Mucinous tumors of the appendix associated with mucinous tumors of the ovary and pseudomyxoma peritonei[colon] a clinicopathological analysis of 22 cases supporting an origin in the appendix. Am J Surg Pathol 1991; 15[colon]415[ndash]29.

Ovary; Carcinoma; Mucinous tumors; Borderline tumors; Epithelial tumors; Microinvasion; Pseudomyxoma peritonei; Pseudomyxoma ovarii; Prognosis; Tumor stage; Appendix

[copy] 2000 Lippincott Williams [amp] Wilkins, Inc.