The reported experience with early-stage (FIGO stage I/II) ovarian carcinoma (OC) is limited given that the majority of women with OC are diagnosed at an advanced stage. There has not been an extensive review of these tumors, and since the pathologic criteria differentiating invasive and borderline tumors have evolved over time, the issue of whether a proportion of these tumors should be reclassified has not been addressed. We identified patients with stage I/II invasive OC who underwent primary surgical management at Memorial Sloan-Kettering Cancer Center from 1980 to 2000. Patients known to have a BRCA mutation or a family history of breast/ovarian cancer were excluded. Hematoxylin and eosin slide review, blinded to clinical outcomes, using current diagnostic criteria for ovarian carcinomas and borderline ovarian tumors, was performed. Progression-free survival (PFS) and disease-specific survival (DSS) were estimated and compared. Hematoxylin and eosin slides were reviewed for 140 of the 145 patients identified. The diagnosis was changed to borderline (low malignant potential) in 41 cases (29.3%). Twenty-nine (70.7%) of 41 changes in diagnosis involved endometrioid and mucinous tumors. This was attributable to the application of recently revised criteria for distinguishing borderline tumors from carcinomas. None of the originally diagnosed clear cell carcinomas was reclassified as borderline. The distribution of histologic subtypes among the 94 carcinomas included 26 serous (27.7%), 25 clear cell (26.6%), 22 endometrioid (23.4%), 10 mixed (10.6%), 6 mucinous (6.4%), 2 malignant Brenner (2.1%), and 3 adenocarcinomas, not otherwise specified (3.2%). Adjuvant therapy was given to 84 (89.4%) of the 94 patients with carcinomas. The 5-year PFS and DSS were significantly greater for the group of cases that was reclassified as borderline (4.5% vs. 26.2% progressed [P = 0.006]; 4.5% vs. 25.6% died [P = 0.003]). The 5-year PFS and DSS were significantly worse for carcinomas with a TP53 mutation (22.6% vs. 41.2% progressed [P = 0.04]; 21.7% vs. 24.7% died [P = 0.04]). There were no statistically significant differences in outcome between stages I versus II, tumor grades, clear cell histology versus other, and stage IC preoperative versus intraoperative rupture. We concluded that a large number of cases originally diagnosed as early-stage sporadic OC were borderline tumors. Clear cell histology does not confer a worse prognosis compared with other histologies. The presence of a TP53 mutation was an adverse prognostic indicator.
Ovarian cancer is the fifth most common malignancy among women, with an estimated 25,400 new cases and 14,300 deaths for 2003.23 It is the second most common gynecologic malignancy, and the most lethal, accounting for more than half of all deaths from gynecologic malignancies.23 Survival controlled for stage is similar to that of other gynecologic malignancies, with favorable survivals reported for early-stage (FIGO stage I/II) ovarian carcinomas.22,37 However, the majority of patients (>60–70%) present at an advanced stage, where the outcome is significantly worse.22,37 Considerable literature exists about the clinical and pathologic characteristics of advanced-stage ovarian carcinoma, but our knowledge regarding early-stage ovarian cancer is somewhat limited. This has led to uncertainty concerning the appropriate management of patients with early-stage disease.
There has been much interest in determining which prognostic factors could help guide the management of patients with early-stage ovarian cancer (OC). Numerous studies have suggested that tumor grade could be an important determinant of outcome in this group, but the importance of other prognostic factors, such as FIGO substage, patient age, tumor histology, ovarian capsular rupture, ascites, tumor growth on the ovarian surface, malignant peritoneal washings, adhesions, and DNA ploidy remains unclear.1,9,16,22,26,33,45,52,53,61,62,68,71
An explanation for this dilemma lies in the heterogeneity of patient populations studied in early-stage OC reports. Some include clinically staged patients as opposed to surgically and pathologically staged patients and cases that have not undergone stringent pathologic review.1,16,33,53,61,62 Young et al66 reported that 31% of patients with disease grossly confined to the ovaries and/or pelvis were upstaged after an adequate surgical staging procedure. Subsequent studies have confirmed the need for multiple peritoneal biopsies, peritoneal washings, omentectomy, and pelvic and para-aortic lymph node sampling (comprehensive surgical staging).13,26,38,45,71 In general, patients who are comprehensively staged and are thought to have low-risk disease (ie, FIGO stage IA or IB, grade 1) have an excellent survival and do not require adjuvant therapy.1,36,45,65,67–69 Therefore, we hypothesized that studying a group of surgically and pathologically staged patients could minimize the importance of some candidate prognostic factors.
Inherent biologic properties of ovarian tumors could also influence outcomes. For example, BRCA-related (“hereditary”) ovarian carcinomas, constituting approximately 10% of ovarian carcinomas,10 appear to have a better outcome compared with non–BRCA-related (“sporadic”) cases.8,11,44 Despite accounting for a small proportion of all OCs, inclusion of such cases in clinicopathologic studies might further contribute to the uncertainty regarding the significance of various prognostic indicators in early-stage disease.
The World Health Organization (WHO) pathologic criteria for distinguishing invasive OCs from borderline tumors were described in 1973 and updated in 1999.48 Differentiating these two entities can be difficult and is subject to significant interpathologist and intrapathologist variation.2 Aside from the controversy concerning whether noninvasive, cytologically low-grade micropapillary and cribriform serous tumors represent carcinomas or borderline tumors,15,17,40,50,56 there have been numerous recent publications that have sharpened the criteria for separating borderline tumors and carcinomas. Among the most important have been those that describe morphologic features that distinguish mucinous borderline tumors and carcinomas from mimics that include metastatic mucinous carcinomas involving the ovary.28,42,51 Other significant contributions include new diagnostic criteria for well-differentiated endometrioid and mucinous carcinomas.6,27,41 Outcome analysis, therefore, would be greatly affected by the criteria defining carcinomas and borderline tumors.
TP53 is a tumor suppressor gene that is involved in cell-cycle control and arrest.30 It is the most commonly mutated gene in almost all malignancies including ovarian cancer.57 Alterations in its protein product, p53, have been evaluated as a prognostic factor in advanced-stage OC, with highly conflicting results.18,19,21,29,32,35,46,54,55,64 The prognostic value of gene mutations, however, has not been subject to extensive analysis. We are not aware of reports regarding its significance in early-stage sporadic OC.
The aims of this study were to review all cases of early-stage sporadic OC, primarily treated at our institution, and to apply current criteria for distinguishing carcinomas from borderline tumors to determine whether changes in classification affected the clinicopathologic features of this group. We also sought to determine relevant prognostic factors, including histopathologic and clinical parameters and the impact of TP53 mutation on prognosis.