Complete lymph node dissection: is it essential for the treatment of borderline epithelial ovarian tumors?

ObjectivesIntraoperative frozen section has greater than 90% accuracy for ovarian tumors; however, mucinous histology has been shown to be associated with increased frozen section inaccuracy. Recent data demonstrate that primary ovarian mucinous carcinomas have no lymph node involvement, even when extraovarian disease is present, and therefore may not require lymph node dissection. Our primary objective is to evaluate the accuracy of identifying mucinous histology on frozen section.Methods/MaterialsA cross-sectional review of mucinous ovarian tumors in surgical patients at one institution from 2006 to 2016 was performed. Cases reporting a mucinous ovarian tumor on frozen section or final pathology were identified. Frozen section results were compared with final diagnosis to calculate concordance rates. Analyses with χ2 and t tests were performed to identify variables associated with pathology discordance.ResultsA total of 126 mucinous ovarian tumors were identified. Of these, 106 were reported as mucinous on frozen section and 103 (97.2%) were concordant on final pathology. Discordant cases included 2 serous and 1 clear cell tumor. Among the 103 mucinous tumors, classification as malignant, borderline, or benign was concordant in 74 (71.8%) of 103 cases, whereas 22 (21.4%) of 103 were discordant and 7 (6.8%) were deferred to final pathology. Lymph node dissection was performed in 33 cases; the only case with lymph node metastasis was a gastrointestinal mucinous adenocarcinoma. Discordance between frozen section and final pathology was associated with larger tumor size and diagnosis other than benign: discordant cases had a mean tumor size of 21.7 cm compared with 14.4 cm for concordant cases (P < 0.001), and 93.5% of discordant cases were borderline or malignant, compared with 30.5% of concordant cases (P < 0.001).ConclusionsIntraoperative identification of mucinous histology by frozen section is reliable with a concordance rate to final pathology of 97.2%. No lymph node metastases were present in any malignant or borderline primary ovarian cases.

Download Full-text

Survival of SLNB-positive melanoma patients with and without complete lymph node dissection: A multicenter, randomized DECOG trial.

Journal of Clinical Oncology ◽

10.1200/jco.2015.33.18_suppl.lba9002 ◽

2015 ◽

Vol 33 (18_suppl) ◽

pp. LBA9002-LBA9002 ◽

Cited By ~ 9

Author(s):

Ulrike Leiter ◽

Rudolf Stadler ◽

Cornelia Mauch ◽

Werner Hohenberger ◽

Norbert Brockmeyer ◽

...

Keyword(s):

Lymph Node ◽

Lymph Node Dissection ◽

Distant Metastases ◽

Tumor Burden ◽

Phase Iii ◽

Node Dissection ◽

Significant Treatment ◽

Melanoma Patients ◽

Complete Lymph Node Dissection

LBA9002 Background: Complete lymph node dissection (CLND) following positive sentinel node biopsy (SLNB) was evaluated in a randomized phase III trial. Methods: 1,258 patients with cutaneous melanoma of the trunk and extremities and with positive SLNB were evaluated. Of these, 483 (39%) agreed to randomization into the clinical trial. 241 patients underwent observation only, 242 received CLND. Both groups had a subsequent 3-years follow-up. Recurrence-free (RFS), distant metastases free (DMFS) and melanoma specific (MSS) survival were analyzed as endpoints. Results: Patient enrolment was performedfrom January 2006 to December 2014. In the intent to treat analysis, both groups did not differ significantly in distribution of age, gender, localization, ulceration, tumor thickness (median 2,4 mm in both groups), number of positive nodes, or tumor burden in the SN. The mean follow-up time was 34 months (SD ± 22.1). No significant treatment-related difference was seen in the 5-years RFS (P = 0.72), DMFS (P= 0 .76) and MSS (P = 0.86) in the overall study population. Conclusions: In this early analysis of trial results, no survival benefit was achieved by CLND in melanoma patients with positive SLNB. A subsequent analysis three years after inclusion of the last patient is planned.

Download Full-text