Restaging Surgery for Women with Borderline Ovarian Tumors: Results of a French Mulitcenter Study
Abstract & Commentary
Synopsis: Women who initially were diagnosed with Stage IA disease and who had serous borderline tumors or underwent cystectomy appeared to derive the most benefit from restaging surgery. Nonetheless, the indications for restaging surgery remain controversial, as no difference in recurrence rate was observed between women who underwent restaging and those who did not.
Source: Fauvet R, et al. Cancer. 2004;100:1145-1151.
Borderline ovarian tumors account for about 10-20% of all ovarian epithelial tumors and are characterized by occurrence in younger women, earlier stage at diagnosis and better prognosis compared to invasive ovarian cancer. In some cases, the diagnosis is made after surgery, where neither complete ovarian resection (conservative operation) nor complete surgical staging information has been obtained, leaving the clinician in a quandary as to the appropriate next step. Fauvet and colleagues attempt to address this clinical situation through a review of all ovarian borderline or low malignant potential (LMP) tumors of the ovary diagnosed in their hospital system over a 12-month period. The principal aims of this retrospective review were to determine if conservative surgery was detrimental, if incomplete staging warrants a restaging operation, and what risk factors are associated with upstaging. Although no centralized pathological review of the individual cases was made, adherence to the current FIGO recommended classification scheme was followed. Three-hundred-sixty women with an ultimate final diagnosis of LMP tumor were accessioned. Of these, 150 or 42% underwent intraoperative histologic examination from which 97 (65%) were correctly identified as having LMP tumors. The remainder were classified as either benign (23%), carcinoma (2%) or not definitive (11%). Using Fauvet et al’s definition of staging (to include peritoneal cytology and biopsies, omentectomy and in the case of mucinous tumors, appendectomy), 37 women (38%) underwent complete evaluation. In this cohort, just 5 were treated conservatively defined as leaving the uterus and at least part of an ovary behind. Among the "staged" group, an additional 25 "high-risk" patients were added, as they underwent the formal staging procedure despite having inconclusive intraoperative histology.
In this cohort, one additional patient was treated conservatively, leaving 6 of 62 patients treated conservatively. Of the 298 incompletely staged or unstaged patients, 54 (18%) underwent a re-staging procedure. Although a priori criteria for re-staging was not discussed or presented, patients undergoing the procedure were significantly younger and more likely to have had an initial conservative procedure. Interestingly, only half of those women undergoing a second procedure actually met Fauvet et al’s criteria of staging. However, the majority of those undergoing a second operation were done so with conservation of fertility as a goal (n = 48, 89%). Of these 54 secondary operations, more advanced disease was identified in 8 patients (15%), with 3 upstaged to stage IB, 1 to stage IIA, 1 to stage IIB, 2 to stage IIIA and 1 to stage IIIC. With the exception of the latter, they were all initially stage IA.
No significant characteristics were associated with upstaging although these tumors were more often serous and in women who had a cystectomy as a primary operation. At a median follow-up of 37 months, overall survival was nearly identical between those undergoing staging operations compared to those who did not. Overall, 34 patients (10%) recurred, the majority of which were those who had undergone conservative operations (25 of 160 total or 16%). This was the only factor determining recurrence risk including whether or not they underwent a re-staging procedure. Fauvet et al’s conclude that little benefit is gained by reoperating on a patient for the purposes of staging unless a patient has a stage IA serous tumor and has undergone cystectomy as a primary operation. However, even in this scenario, no survival benefit is gained by the maneuver and its conduct remains controversial.
Comment by Robert L. Coleman, MD
The results of this retrospective study of tumors fits nicely into the growing body of literature outlining the natural history of LMP ovarian tumors. Fortunately, long-term survival for women diagnosed with this disease is the rule rather than the exception and outside of those uncommon tumors demonstrating metastatic disease at diagnosis and those with invasive implants, these patients remain predominately relapse-free. As confirmed in this review, those undergoing conservative procedures, that is, in whom ovarian tissue is left behind, are the ones most at risk for relapse. However, the majority of these recurrences are of similar histology and are reliably salvaged with additional surgery to remove the adnexa. Fertility sparing should be considered in those who are interested in childbearing. These findings should be of some comfort to practitioners and patients alike.
However, it is important not to equate these survival characteristics with benignity. For instance, it is remarkable that in the current series more than 58% of the 360 LMP tumors accessioned did not have an intraoperative histological assessment (frozen section). One must remember that patients may not only recur with invasive disease but also can succumb to its "benign" natural history through indolent and progressive growth. In addition, since re-staging, as defined in this report, is not associated with variable survival dynamics, the procedure may errantly be omitted under the guise that it won’t make a difference. While little appears to be gained in those patients with no macroscopic disease by a secondary staging procedure, it shouldn’t preclude staging when the diagnosis is suspected intraoperatively.
As has been reported elsewhere as well as in this study, the final and intraoperative histological diagnoses are non-correlative in a significant fraction (> 20%).1-3 Absence of important intraoperative information makes subsequent treatment decisions more problematic and may lead to both under- and overtreatment. This was highlighted in a recent review of 2 hospital systems where surgical staging for suspected early ovarian cancer was dichotomized by the presence of a gynecological oncologist. Where staging was not done routinely and treatment decisions were made on the basis of postoperative ovarian histology, approximately 20% more patients were treated than when operative information correctly identified the patients’ stage. In addition recurrence rate was reduced from 28% to 10% by correctly identifying true "at-risk" patients. It is unlikely a similar policy for LMP tumors will have such a dramatic effect but we are nonetheless, left to make the decision to stage or not intraoperatively, and until we can be confident that our final histology will be represented by frozen section we must address the situation conservatively. It is likely we will increase our diagnostic precision for LMP, allowing us to be more selective in whom we stage. Once the diagnosis is confirmed, limited stage patients appear to gain very little from addition surgical evaluation.
1. Querleu D, et al. Br
J Obstet Gynecol. 2003;110: 201-204.
2. Lin PS, et al. Cancer. 1999;85:905-911.
3. Winter WE, et al. Obstet Gynecol. 2002;100:671-676.
Robert L. Coleman, MD, Dept. of Gynecologic Oncology, University of Texas Southwestern Medical Center, Dallas, TX, is Associate Editor for OB/GYN Clinical Alert.